design and implementation of eight-legged robotic transporter

DESIGN AND IMPLEMENTATION OF EIGHT-LEGGED ROBOTIC

TRANSPORTER

A Thesis

presented to

the Faculty of California Polytechnic State University,

San Luis Obispo

In Partial Fulfillment

of the Requirements for the Degree

Master of Science in Mechanical Engineering

by

Jeremy David DePangher

September 2013

ii

© 2013


ALL RIGHTS RESERVED

iii

COMMITTEE MEMBERSHIP

TITLE: Design and Implementation of Eight-Legged Robotic

Transporter

AUTHOR: Jeremy David DePangher

DATE SUBMITTED: September 2013

COMMITTEE CHAIR: Saeed Niku, PhD

Professor of Mechanical Engineering

COMMITTEE MEMBER: John Ridgely, PhD


COMMITTEE MEMBER: Peter Schuster, PhD


iv

ABSTRACT

Design and Implementation of Eight-Legged Robotic Transporter


This thesis contains the design, manufacturing, and testing of a functional eight-

legged robotic transporter based on the concept design laid out in U.S. Patent 7,246,671.

The device is intended to achieve three different sequences of motion: regular driving,

obstacle climbing, and stair climbing. The prototype was carried through concept design,

analysis, selection of materials and components, manufacturing, software development,

and final assembly and testing. The device can be assembled under multiple

configurations, which harbor certain advantages and disadvantages. The results of the

testing encourage the continuation of a second iteration of this concept.

Keywords: U.S. Patent 7,246,671, functional, eight-legged, robotic, obstacle, stair,

climbing, design, testing, sequence, prototype, configurations.

v

ACKNOWLEDGMENTS

I would like thank my committee for their continued support of my project: Dr.

Saeed Niku, my committee chair; Dr. John Ridgely; and Dr. Peter Schuster. These

individuals aided me each step along the way, and for that I am very grateful.

My project would not have been possible without the financial support and

encouragement of my sponsor, Mr. Michael Goren.

I would also like to acknowledge the technicians at the Mustang ’60 machine

shop for their help in manufacturing many of the parts necessary for the completion of

this project.

Lastly, I would like to thank my family and friends for the motivation to see this

project to completion.

vi

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................... viii

LIST OF FIGURES ........................................................................................................... ix

CHAPTER

1. INTRODUCTION ........................................................................................................ 1

1.1. Background ............................................................................................................. 1

1.2. Patent....................................................................................................................... 1

1.3. Scope and Objectives .............................................................................................. 2

2. DESIGN AND ANALYSIS ......................................................................................... 5

2.1. Concept ................................................................................................................... 5

2.1.1. Independent Joint Motions ................................................................................ 5

2.1.2. Linear Actuation ............................................................................................... 5

2.1.3. Steering and Driving ......................................................................................... 6

2.1.4. Moving Sequences ............................................................................................ 9

2.2. Mechanical Design................................................................................................ 22

2.2.1. Configuration Options .................................................................................... 23

2.2.2. Analysis........................................................................................................... 29

2.2.3. Component Selection ...................................................................................... 44

2.3. Mechatronics Design ............................................................................................ 51

2.3.1. Linear Actuator Control .................................................................................. 51

2.3.2. Driving and Steering Motor Control ............................................................... 53

2.3.3. Microprocessor Selection................................................................................ 55

2.3.4. Power Supplies................................................................................................ 56

2.3.5. Overall System Configuration ........................................................................ 58

2.3.6. Software .......................................................................................................... 59

2.4. Manufacturing ....................................................................................................... 61

3. TESTING .................................................................................................................... 64

3.1. Plan and Objectives............................................................................................... 64

3.2. Stair Climbing Sequence Tests ............................................................................. 64

vii

3.3. Obstacle Climbing Sequence Tests ....................................................................... 69

4. CONCLUSIONS AND RECOMMENDATIONS ..................................................... 74

BIBLIOGRAPHY ............................................................................................................. 81

APPENDICIES

A: PART DRAWINGS .................................................................................................... 82

B: ENGINEERING CALCULATIONS .......................................................................... 99

C: BILL OF MATERIALS ............................................................................................ 111

D: DATA SHEETS ........................................................................................................ 114

E: SEQUENCE TESTING IMAGES ............................................................................ 117

F: SOFTWARE CODE ................................................................................................. 122

viii

LIST OF TABLES

2.1. Configuration Options for Driving and Steering Motor Pairs ..................................... 7

2.2. Standard Stair Step Sizes ........................................................................................... 13

2.3. Summary of Stair Climbing Sequence ....................................................................... 22

2.4. Variable Inputs for Driving Torque Calculations ...................................................... 31

2.5. Inputs for Drive Motor Mounting Bracket Deflection ............................................... 34

2.6. Inputs for Deflection of Wheel Shaft ......................................................................... 35

2.7. Inputs for Tension and Reduction of Timing Belt ..................................................... 36

2.8. Inputs for Steering Motor Torque .............................................................................. 38

2.9. Inputs for Deflection of Steering Motor Mounting Plate .......................................... 39

2.10. Inputs for Horizontal Support Shaft Deflection ....................................................... 40

2.11. Inputs for Wheel Shaft Press Fit .............................................................................. 41

2.12. Inputs for Actuator Mounting Screw Stress............................................................. 42

2.13. Inputs for Bending Stress in Platform ...................................................................... 43

2.14. Arduino Mega 2560 Pin Usage Breakdown ............................................................ 55

2.15. Current Draw Requirements .................................................................................... 57

2.16. Current Draw for Various Conditions ...................................................................... 57

2.17. Global Variable Declarations ................................................................................... 60

2.18. Function Properties and Descriptions ...................................................................... 61

C.1. Bill of Materials ...................................................................................................... 111

ix

LIST OF FIGURES

1.1 - Concept Sketch Based on Patent Description............................................................. 4

2.1 - Conceptual Platform with Labeled Pairs .................................................................... 8

2.2 - Regular Driving Mode 1 ........................................................................................... 10

2.3 - Regular Driving Mode 2 ........................................................................................... 10

2.4 - Obstacle Climbing Sequence Stage 1 ....................................................................... 11


2.6 - Obstacle Climbing Sequence Stage 3a ..................................................................... 12

2.7 - Obstacle Climbing Sequence Stage 3b ..................................................................... 12


2.9 - Device with Critical Climbing Dimensions.............................................................. 14

2.10 - Stage 1 of Climbing Sequence................................................................................ 15









2.19 - Stage 10 of Climbing Sequence.............................................................................. 20

2.20 - Stage 11 of Climbing Sequence.............................................................................. 21

2.21 - Standard Ball Thrust Bearing ................................................................................. 24

2.22 - Banded Ball Thrust Bearing ................................................................................... 24

2.23 - Turntable Bearing with Assembly .......................................................................... 25

2.24 - Flex Coupling Connection ...................................................................................... 27

2.25 - Wheel Housing with Bearing Pillow Blocks and Wheel ........................................ 28

2.26 - Actuator Tip Being Inserted Into Block to Complete Assembly............................ 29

2.27 - Diagram of Driving Motor Bracket Analysis ......................................................... 33

2.28 - Contact Patch of Drive Wheel ................................................................................ 37

2.29 - Right Angle Geared Motor for Driving .................................................................. 45

2.30 - Inline Geared Motor for Steering ........................................................................... 46

2.31 - HCR112-14-12 Flex Coupling ............................................................................... 48

2.32 - Turntable Bearing ................................................................................................... 48

2.33 - Firgelli Automation FA-400-L-12-30 Linear Actuator .......................................... 49

2.34 - Linear Bearing ........................................................................................................ 50

2.35 - Pololu VNH2SP30 Dual Motor Driver Board ........................................................ 53

x

2.36 - Pololu VNH5019 Single Motor Driver Board ........................................................ 54

2.37 - Arduino Mega 2560 ................................................................................................ 55

2.38 - Electronic Components Assembly Diagram........................................................... 58

2.39 - Frame Base with Opposite-Facing Actuator Slots ................................................. 62

2.40 - Actuator Support..................................................................................................... 63

3.1 - Stage 1 to Stage 2 Transition .................................................................................... 65

3.2 - Completion of Stage 2 .............................................................................................. 66


3.4 - Completion of Stage 10 ............................................................................................ 68

3.5 - Rigid Rod Added to Reduce Tilting ......................................................................... 69

3.6 - Vehicle in Position to Begin Stage 1 ........................................................................ 70



4.1 - Complete Solid Model of 8-Legged Robotic Transporter ........................................ 74

4.2 - Complete 8-Legged Robotic Transporter ................................................................. 75

E.1 - Stair Climbing Stage 1 ........................................................................................... 117









E.10 - Stair Climbing Stage 10 ....................................................................................... 121

1

1. INTRODUCTION

1.1. Background

This thesis is concerned with the designing, manufacturing, and testing of a

functional prototype based on the concept design laid out in U.S. Patent No. 7,246,671.

This patent, entitled “Stair-climbing human transporter,” is the sole inspiration for the

body of work done in this thesis.

1.2. Patent

According to the patent language, the purpose of this invention is to “transport

persons, robots or cargo, moving on a surface, over obstacles and for ascending and

descending stairs” [1]. There are several aspects of this invention that not only make it

unique from other transportation methods, but also advantageous in comparison.

First, “typically such vehicles cannot climb stairs while keeping their platform

and rider level” [1]. Perhaps the most critical example of this is stairs. There currently is

not a device that can travel upstairs while keeping its platform flat and also

simultaneously meeting the next two requirements.

Second, a main advantage is that the device can transport persons or cargo with

multiple different configurations of motion.

“while there are many vehicles that can transport multiple individuals and a payload,

many such vehicles are not lightweight and capable of navigating or climbing over

obstacles, climbing stairs and also of transporting more than one individual rapidly on

common surfaces such as roads and sidewalks.” [1]

2

This language points out two key factors at issue in current designs: weight of the

vehicle, and having multiple modes of motion to accomplish climbing over obstacles as

well as travel on flat ground with reasonable speed.

Third, most vehicles face some level of setback when obstacles are in their path.

“Most common vehicles such as automobiles, motorcycles, wheelchairs and gurneys,

place all of their points of ground contacts such as wheels on the ground all at the same

time, such that when there is an obstacle in the road, such vehicles must run over such

obstacle. Often this results in passengers either feeling bumps or having the entire

platform or chassis incline rather than staying level.” [1]

Being able to avoid obstacles entirely without having to run over them or tilt the

platform to overcome them are desirable qualities that common transportation vehicles do

not have.

These three factors are the most important features that were considered for this

project. Each of these was carefully considered and later implemented in the design and

creation of the device.

Under the Claims section of the patent, various descriptions are given of the

components and how they interact with each other. While the scope of this project does

not include every claim specified in the patent, each of the components that were

designed in order to create a functional prototype were based on these claims.

1.3. Scope and Objectives

The scope of this thesis consists of extracting the description of the device’s

components and motions from the patent language, and creating a physical design based

on these ideas. The following is a list of characteristics that must be included in the

design for the functional prototype. The prototype must:

3

Be able to travel up a flight of stairs.

Have eight vertical legs, four of which can move in a plane parallel to the ground

via horizontal actuation of some sort. The other four must remain fixed to the

platform.

Have a platform upon which a weight can be rested to simulate a person or cargo.

Be powered from its own on-board power supply.

Stay level at all times.

For clarification, there were several features included in the patent which did not have to

be represented in this first prototype. The prototype did not have to:

Use sensors.

Be wirelessly controlled.

Be autonomous.

Be made out of a specific type of material.

Using these specifications, a concept design could be formulated. The design

phase, including calculations, selection of materials and components, and any necessary

analysis, follows from the concept design. Once the critical portions of the design were

decided, the relevant parts could be machined. The machined parts would be quality

checked and assembled along with the electronic components. Once the electronics are

fully integrated with the mechanical assembly, testing of the device’s functions could

commence. The testing is to ensure that not only the calculations and analytical

predictions of the design were adequate, but also that the device met the specifications.

Once it is verified that the prototype meets all the specifications, the project is complete.

4

The final version of the prototype will look and function something like the

diagram in Figure 1.1.

Figure 1.1 - Concept Sketch Based on Patent Description

Figure 1.1 shows that the prototype will have eight vertical legs which can each

extend and retract. Four of these vertical legs will be attached to horizontal actuators that

can extend and retract the vertical legs relative to the platform. Each vertical leg will have

a wheel at its base, which can rotate in either a free or controlled manner. Each of the

vertical legs must also be able to steer the wheel either via free or controlled motion. The

platform at the center of the device will be flat and be able to carry cargo to simulate a

load.

5

2. DESIGN AND ANALYSIS

2.1. Concept

The general concept for the type of motion and types of components used to make

that motion happen are described in the patent itself. However, in order to turn the

general patent language into something that could function mechanically, there were

several additional factors that had to be considered. These factors include:

Determining the minimum number of independent joint motions

How to accomplish the linear actuation

How to accomplish the steering and driving of each of the vertical legs

2.1.1. Independent Joint Motions

There are twelve linear actuators – eight vertical and four horizontal. Each of

these constitutes one potential joint of motion. At the tip of four of the vertical legs will

be assemblies containing a steering and a driving motor.1 This constitutes an additional

eight joints of motion for a total of twenty.

2.1.2. Linear Actuation

A component was needed to accomplish the raising and lowering of all eight

vertical legs, as well as the extension and retraction of the fore and aft pairs of vertical

legs relative to the platform. The type of linear actuation could be narrowed down to two

choices. Either a fully assembled linear actuator component could be purchased, or a

linear actuator could be built from smaller components. Due to time and manufacturing

1 Reasons for only four of the legs containing motorized steering and driving are discussed 2.1.3.

2 Discussion of the specific choice of the type of actuator is continued in 2.2.

6

constraints, the former option was chosen. In this way, the eight vertical legs in addition

to the four horizontal actuators would each be controlled by a linear actuator,

electronically controlled by the microprocessor unit.2

2.1.3. Steering and Driving

The choice of having four pairs of vertical legs with steering and driving motors is

not arbitrary. There are several possible configurations for which vertical legs have

steering and driving motors. Theoretically, the number configurations of steering and

driving motors could range all the way from all four pairs having steering and driving

motors to just one pair having them. In addition to the number of pairs of legs with

motors, any combination of configurations is also possible. For example, the front and

back vertical pairs could have driving and steering motors while the middle two pairs do

not. All of these considerations result in a large number of possible configurations, not all

of which are necessary to consider for the actual design. An approach was undertaken to

reduce the number of configurations that would need to be tested. The number and

configuration of pairs of legs with motors is determined by the required functionality of

the prototype. Since the prototype must accomplish regular driving on flat surfaces,

climbing over obstacles, and climbing up stairs, the pairs must be chosen in order to

satisfy each of these challenges.

In the case of regular driving on flat surfaces, the platform could be modeled like

a two-wheel drive car. Regular turning and driving can be accomplished with

2 Discussion of the specific choice of the type of actuator is continued in 2.2.

7

motors on just one pair of legs. Since this would mean that the other pair

contacting the ground would be free, the motorized pair would need to be in the

front if driving forward, and in the back if driving backwards. Caster wheels,

which are free in both driving and steering, could not lead the motion.

Climbing over obstacles would require at most motors on two pairs of legs. This

is because there is never an instance when there would be more than two pairs of

legs not touching the ground. Given all combinations of pairs that are contacting

or not contacting the ground at any given time, only two pairs of vertical legs

would need steering and driving motors.

Climbing up stairs is a more complicated process that implements all the available

joints of motion in the system. But even after analysis of each of the stages of

motion, at any given time at most only two pairs of legs would be utilizing their

steering and driving motors to move the device.3

These considerations allow the design to be simplified to only six possible

configurations involving motors on only two pairs of legs. Table 2.1 shows all possible

configurations. Figure 2.1 labels each pair on the platform for reference.

Table 2.1. Configuration Options for Driving and Steering Motor Pairs

Configuration Pair 1 (Fore) Pair 2 Pair 3 Pair 4 (Aft)

1 - -

2 - -

3 - -

4 - -

5 - -

6 - -

3 See 2.1.4.

8

Figure 2.1 - Conceptual Platform with Labeled Pairs

The best configuration is ultimately determined by testing. However, basic

analysis reveals that configurations 3 and 4 are insufficient. During a regular driving

sequence on flat ground, the default would be to have pairs 2 and 3 on the ground with

pairs 1 and 4 lifted up.4 In order for the device to move at all, at least one of Pairs 2 or 3

must have motors. This eliminates configuration 3. While climbing over obstacles, there

are foreseeable instances when 2 and 3 must be lifted up to traverse the obstacle, and 1 or

4 must drive forward or reverse in order to put 2 and 3 back on the ground. This

eliminates Configuration 4. Configurations 1, 2, 5, and 6 remain as possible

configurations with which the device can operate.

Even though these four remaining configurations could work in theory, further

inspection shows that Configurations 1 and 6 create unnecessary disadvantages. During

4 Having Pairs 2 and 3 on the ground during regular driving would provide the most balance in the system,

since these pairs are only creating transverse loads, not transverse and bending loads.

9

the stair climbing sequence, there would be times where the two front pairs are on the

stairs and the two back pairs are on the ground. One of the front pairs and one of the back

pairs should have motorized wheels that can brake in place to keep the device from

slipping. Considering this as a safety precaution, Configurations 1 and 6 are not ideal.

This leaves Configurations 2 and 5 to test. Testing each of these remaining configurations

would ultimately determine which one is optimized for all conditions.5 In order to test

these two configurations, however, the legs would have to be designed so that the

components could be easily interchanged.

2.1.4. Moving Sequences

As previously mentioned, the three major applications of this device will be to 1)

travel on flat ground, 2) climb over obstacles, 3) climb up and down stairs. The most

frequent usage of the vehicle will be on flat ground, and this application is discussed first.

Climbing over obstacles and climbing up and down stairs are secondary applications that

will be discussed next.

The primary application of the device will be to travel on flat ground. When the

vehicle is travelling this way, there are two likely modes of operation. The first one has

Pairs 2 and 3 on the ground while Pairs 1 and 4 are raised. In order for this mode to be

feasible as a regular driving mode, a motorized pair would have to be put on Pair 2. This

configuration is named Mode 1. It is sketched in Figure 2.2.

5 See 3 for a determination of the final configuration.

10

Figure 2.2 - Regular Driving Mode 1

The second mode of operation has Pairs 1 and 4 on the ground while Pairs 2 and 3

are raised. This mode, Mode 2, is less common because during regular driving there is no

necessity for Pairs 1 and 4 to be controlling the motion and therefore inducing

unnecessary bending in their horizontal actuators.6 Mode 2 is shown in Figure 2.3.

Figure 2.3 - Regular Driving Mode 2

The next major application of the device is to traverse obstacles. A brief sequence

is laid out in this section to demonstrate how this will be accomplished. Stage 1 involves

extending Pair 1 over the obstacle, while keeping the rest of the pairs in place. Figure 2.4

shows Stage 1.7

6 Even if the horizontal actuators are completely retracted, the fact that Pairs 1 and 4 are farther away than

Pairs 2 and 3 from the center of gravity means that there will be additional stress on the platform. If this can

be avoided, it should be. 7 All arrows in Figures 2.3 – 2.18 indicate the direction of motion that took place in order for the joint to

arrive at that position. For example, an up arrow on a vertical actuator means that the actuator retracted to

11

Figure 2.4 - Obstacle Climbing Sequence Stage 1

Once Pair 1 is extended over the obstacle, Pairs 1 and 4 must be lowered to the

ground so that Pairs 2 and 3 can traverse the space over the obstacle. Stage 2 is shown in

Figure 2.5.


Stage 3 involves carrying the platform over the obstacle. This can most easily be

accomplished by rolling Pairs 1 and 4 forward as far as possible. If necessary, depending

upon the size of the obstacle, extension of the horizontal actuators might be required for

extra distance. Stage 3 is shown in Figure 2.6.

get to that position. It does not indicate the motion of the next stage. Labels such as “P1” refer to the pair of

actuators, such as “Pair 1,” being discussed in the text.

12

Figure 2.6 - Obstacle Climbing Sequence Stage 3a

Figure 2.6 shows that in order for the platform to completely traverse the obstacle,

Pair 4’s horizontal actuators must extend past the obstacle. Figure 2.7 demonstrates what

is, in this case, a necessary second part to Stage 3.

Figure 2.7 - Obstacle Climbing Sequence Stage 3b

In order to complete the sequence, Pairs 2 and 3 must lower to the ground and

Pairs 1 and 4 must return to their original raised position. Figure 2.8 shows Stage 4, the

final stage. After Stage 4, the device will begin driving on flat ground again.

13


The final application of the device is to climb up a full flight of stairs. Because

stairs generally have certain standard dimensions, the platform would need to be designed

in a way that would allow it to climb any set of stairs while keeping the platform itself

level with the ground. Using average stair riser and tread sizes, the dimensions for the

length of the platform as well as the actuator stroke lengths could be determined. Table

2.2 shows these average lengths.

Table 2.2. Standard Stair Step Sizes

Dimension (in)

Riser 8

Tread 11

The length of each linear actuator is determined by its stroke length. If a larger

stroke length is desired, then a longer actuator is required. In order for the actuators to be

in-line with each other and not overhang the edge of the platform, the actuators had to be

kept at a maximum stroke length of 18”.8 This meant that the horizontal pair could extend

to the second step. Because of the large increase in cost for actuators of stroke lengths

8 As compared to the opposing horizontal actuators being offset or placed on top of each other.

14

over 30”, it was decided early on that the prototype would be demonstrated using a

specially fabricated staircase with the same length tread but half-scaled risers. As will be

described in the following diagrams, this riser height of 4” would allow the maximum

height dimension (28”) to be reached with a 30” stroke linear actuator.

Analysis of each stage of the motion of the device up a flight of stairs was the key

to determining certain dimensions. Figure 2.10 through Figure 2.20 describe the eleven

stage process involved in the device travelling from flat ground to being completely on a

set of stairs.

There are a few dimensions that are important to consider for all of the following

diagrams. First, the length of the platform is 48”. Second, there are 5” between Pair 2 and

its closest edge, and Pair 3 and its closest edge. Third, the wheels and their housings at

the base of each leg are approximately 10” tall, which makes the potential height of the

platform 30” + 10” ≈ 40”. Figure 2.9 shows these dimensions in a two-dimensional sketch

of the platform.

Figure 2.9 - Device with Critical Climbing Dimensions

Stage 1

15

Figure 2.10 - Stage 1 of Climbing Sequence

The first stage of motion involves moving Pair 1 from the ground to the second

stair.9 This involves retracting Pair 1’s vertical actuators and extending its horizontal

ones. The maximum distance between Pair 1 and Pair 2 is 23”; 18” of stroke plus 5” from

Pair 2 to the edge of the platform where the stroke begins. To position Pair 1 at the center

of stair 2, the actuator must extend a minimum of 13.5”.10

Stage 2


9 The first ten diagrams show Pair 4 touching the ground. While this is not necessary until Stage 8, it is

shown simply for consistency. Lowering Pair 4 in preparation for a stair climbing sequence would also save

time, since Pair 4 must contact the ground during the sequence. 10

This minimum distance assumes the wheels of Pair 2 are touching the base of the first step. The distance

between the edge of the wheels on Pair 2 and the edge of the platform is 3”. This is overhang over the first

step, so by subtracting this from one and a half times the tread length (16.5”) gives 13.5”.

16

The transfer of Pair 2 from the ground to stair 1 requires three motions. First, Pair

2 must retract. Second, the horizontal actuators connected to Pair 1 must also retract.11

Third, the wheels on Pairs 3 and 4 must roll forward.

Stage 3


Stage 3 is essentially Stage 1 repeated. Pair 1 must move from stair 2 to stair 3 to

clear the way for Pair 2 to move from stair 1 to stair 2. At this point, the only motion

required to achieve Stage 3 is the retraction of the vertical actuators of Pair 1 while also

extending the horizontal actuators of Pair 1.

Stage 4


11 A third motion is required if Pairs 3 and 4 do not have caster wheels. They must drive forward

simultaneous with Pair 2’s horizontal retraction in order to keep Pairs 3 and 4’s wheels from dragging on

the ground.

17

Stage 4 is closely related to Stage 2. Pair 2 must move from stair 1 to stair 2. This

is accomplished by the retraction of Pair 2 and the retraction of Pair 1’s horizontal

actuators.12

The wheels on Pairs 3 and 4 will bring the backside of the platform forward.

Stage 5


Stage 5 is similar to Stage 3. Pair 1 must again move up one stair – this time from

3 to 4. This is accomplished by retracting Pair 1’s vertical actuators and extending its

horizontal ones. While Pair 1’s vertical actuators have almost retracted to zero position, it

hasn’t quite reached zero which means that Pairs 3 and 4 do not have to extend their

vertical actuators during this stage.

Stage 6

12 The same issue with the caster wheels also arises in this case. See Note 7.

18


Stage 6 is similar to Stage 4. Pair 2 must move from stair 2 to stair 3. This is

accomplished again by retracting Pair 3, and retracting Pair 4’s horizontal actuators. The

wheels on Pairs 3 and 4 must again roll forward.

Stage 7


Stage 7 is similar to Stage 5 but there are now a couple of key differences. Pair 1

must move to the next step, but in order to accomplish this two additional motions must

occur. The vertical legs of Pairs 2, 3, and 4 must all extend to account for the additional

4” for Pair 1 to get to stair 5. At this point the vertical actuators for Pairs 3 and 4 are

extended approximately 14”.

Stage 8

19


Stage 8 begins the pattern that will occur for the remainder of the flight of stairs.

Instead of just moving Pair 2 up one stair, both Pair 2 and Pair 3 will move up one stair

simultaneously. This is accomplished by retracting both Pairs 2 and 3 to fit onto the next

stair. In order to position them horizontally, however, Pair 1’s horizontal actuators must

retract, thus bringing the platform forward. It is also important to note that since there are

38” between Pairs 2 and 3, their vertical legs will no longer be able to be positioned at the

center of the stair. This is acceptable for two reasons. First, the contact patches (and also

the center of gravity) of the wheels are completely on the stair. Second, Pairs 1 and 4

would provide support for the vehicle even if Pair 3 were to slip off its step.

Stage 9

20


Stage 9 is almost identical to Stage 7. Pair 1 must move its leg up one stair. But

again Pairs 2, 3 and 4 must extend their vertical actuators to account for this additional 4”

by Pair 1 moving up a stair.

Stage 10


Stage 10 is similar to Stage 8. Both Pairs 2 and 3 will move up 1 step

simultaneously. This is achieved by Pairs 2 and 3 retracting 4” while Pair 1 retracts its

horizontal actuator to move the platform forward. Pair 4’s horizontal actuators will likely

have to extend slightly at this point. Pair 4 has reached ymax = 28”. At every even stage

from this point forward, Pair 4 must extend to ymax.

21

Stage 11


Stage 11 now combines the motion of both Pairs 1 and 4. This can be

accomplished because the platform is now fully on the stairs, making its pattern of

motion repeatable every 2 stages. Pair 2 and 3 will extend, while Pair 4’s horizontal

actuators retract to put it on the new stair. Simultaneously Pair 1 will extend its horizontal

actuators to place it on the 7th

stair.

Now that the platform is entirely on the stairs, Stages 10 and 11 will be repeated

until the platform reaches the top of the stairs. At this point, Pairs 3 and 4 will do the

inverse of what Pairs 1 and 2 did for the first several stages. With Pairs 1 and 2 on the top

of the staircase, Pairs 3 and 4 must simply alternate one stair at a time until they have also

reached the top.

22

In summary, each stage requires specific joints to move to locate the device at the

correct position. Table 2.3 illustrates the results of each stage.

Table 2.3. Summary of Stair Climbing Sequence

Stage Pair 1 Pair 2 Pair 3 Pair 4

1 Stair 2 Ground Ground Ground

2 Stair 2 Stair 1 Ground Ground






8 Stair 5 Stair 4 Stair 1 Ground



11 Stair 7 Stair 5 Stair 2 Stair 1

In order to travel down a flight of stairs, the sequence must simply do the reverse

of what it did to climb up the stairs. This time, instead of Pair 1 retracting its vertical

actuators and extending its horizontal ones, it will extend its vertical actuators and extend

its horizontal ones. The device can then sequentially lower Pair 2 to reach the top step,

just as Pair 2 was raised to reach the first step in the ascension sequence. The pattern

continues until the platform has reached the bottom of the stairs.

2.2. Mechanical Design

There were four stages involved in the mechanical design of the device. First, the

conceptual design had to be broken down into parts and assemblies that could accomplish

the required functionality. Second, analysis was conducted to prove that each of these

designed parts would be feasible and operate correctly. Third, components were selected

23

based on the analysis. Fourth, a solid model was constructed which could be used to

create drawings for the manufacturing phase.

2.2.1. Configuration Options

The device can be broken down into two main categories: the vertical legs and the

frame. There are two different versions of the vertical legs – one with drive and steering

motors and the other without either. The more complex version contains the drive and

steering motors, and therefore required the most careful analysis and consideration. The

complex vertical leg had to accomplish two goals: 1) rotate the base of the leg assembly

relative to the actuator tip, and 2) rotate the wheel in contact with the ground.

1) The rotation of the leg base could be accomplished by installing a bearing.

The two types of bearings that were considered were turntable bearings (e.g. a

lazy susan bearing) and banded ball thrust bearings. Choosing a ball bearing

would have increased the complexity of the design without adding any

comparative benefits. The bearing needed to survive a thrust load, because the

leg itself would not always be in contact with the ground. Without the band to

keep the inner race and outer race from separating, a regular ball bearing

would have fallen apart the moment the leg lifted off the ground. Figure 2.21

shows a section view of the standard ball thrust bearing, which demonstrates

the lack of feasibility without the band. Figure 2.22 shows a generic banded

ball thrust bearing.

24

Figure 2.21 - Standard Ball Thrust Bearing13

Figure 2.22 - Banded Ball Thrust Bearing14

The main disadvantage of the banded ball thrust bearing is the need for a press

fit to assemble the components together. This would require circular

components to fit inside the inner race, which limits the flexibility for

selecting the size and shape of materials in the assembly. While the banded

ball thrust bearing has disadvantages for this application, a turntable bearing

13 “Thrust Ball Bearings”. NSK Products. http://www.nskeurope.de/cps/rde/xchg/eu_de/hs.xsl/thrust-ball-

bearings.html. 03 Aug. 2013. 14

“Banded Assembly Ball Thrust Bearing”. Reid Supply Company. http://www.reidsupply.com/sku/INA-

430/. 03 Aug. 2013.

http://www.nskeurope.de/cps/rde/xchg/eu_de/hs.xsl/thrust-ball-bearings.html

http://www.nskeurope.de/cps/rde/xchg/eu_de/hs.xsl/thrust-ball-bearings.html

http://www.reidsupply.com/sku/INA-430/

http://www.reidsupply.com/sku/INA-430/

25

could survive in both tension and compression and has the advantage of being

low cost. Another key advantage of the turntable bearing is that it has

convenient locations for attaching the other parts of the assembly. Ultimately,

the turntable bearing was chosen because it could handle the loading

conditions of the leg, was low cost, and could be easily integrated into the

assembly. Figure 2.23 shows how the turntable bearing operates within the

assembly.

Figure 2.23 - Turntable Bearing with Assembly

2) The rotation of the wheel could be accomplished by either using a direct

connection between the motor shaft and the wheel, or through a power train.

Each has a key disadvantage at the outset. Direct attachment requires that the

motor spin speed and torque are the correct amount without being able to

adjust with a gear reduction or some other power train method. The power

train option is more complex and thus requires more components. Because

26

cost was a consideration for every aspect of the design and manufacturing, a

solution had to be found that could allow for the selection of low cost motors.

Eventually cost drove the decision for the direct attachment method.15

The selection of direct attachment for the drive mechanism requires a bracket to

attach the motor to the rest of the leg base. This could be accomplished with a U-shaped

metal bracket with the motor attached on one side and the leg attached on the other. With

direct attachment, care had to be taken to ensure that the motor shaft and wheel shaft

were correctly aligned. Ensuring alignment via manual machining methods only would

be very difficult, so the decision was made to install a flex-coupling between the two

shafts. This would eliminate the negative effects of any small amount of parallel

misalignment between the two shafts. Figure 2.24 shows the motor, U-bracket, wheel

shaft, and flex coupling portion of the assembly.

15 See 2.2.3.

27

Figure 2.24 - Flex Coupling Connection

The wheel shaft was attached to bearings on each end. Banded ball thrust bearings

were selected because, during operation, the shaft could experience forces parallel to the

axis of rotation. As discussed before, a regular ball bearing cannot protect against a thrust

load. These bearings were installed inside their own pillow blocks, which were fastened

to a wheel housing. The wheel housing needed to be shaped in such a way as to allow the

wheel to fit inside while simultaneously allowing for easy fastening to each of the pillow

blocks, as well as the rest of the assembly above it. Figure 2.25 shows how the selection

of square tubing made sense to meet these requirements.

28

Figure 2.25 - Wheel Housing with Bearing Pillow Blocks and Wheel

Figure 2.25 also shows a cap attached on top of the wheel housing to allow for

attachment to the steering drive mechanism (Figure 2.23) on the upper part of the leg.

The last configuration decision was how to attach the actuator tip to the assembly

below it. At this point in the design several different linear actuator types had been

examined, and one with a circular tip had been selected.16 This meant that there needed to

be a part that could attach to the tip but also attach to the plate connected to the turntable

bearing below. The simplest design to accomplish this was a square aluminum rod with a

circular slot on top. There would also be a hole running through the side of the rod that a

16 See 2.2.3.

29

bolt could fit through. The actuator tip came with a 12 mm hole to allow for this sort of

fastening. Figure 2.26 shows how the actuator tip is integrated with the leg base.

Figure 2.26 - Actuator Tip Being Inserted Into Block to Complete Assembly

If the tube inside the actuator does not rotate, then the steering motor would stay

in place as it caused the pulley to rotate the drive mechanism below it. In this way, the

only rotating components would be everything below the steering motor mounting plate.

2.2.2. Analysis

Engineering analysis was conducted for components that were directly related to

certain specifications or needed to meet some threshold in order for the device to

function. The mechanical design analysis includes:

1) Drive motor torque

2) Deflection of drive bracket

30

3) Deflection of wheel shaft

4) Tension and reduction of timing belt

5) Steer motor torque

6) Deflection of steering plate

7) Horizontal support shaft deflection

8) Allowable torque of wheel shaft press fit

9) Tension in linear actuator mounting screws

10) Bending stress and deflection of platform

While all of these calculations are included in Appendix B, there were several

considerations and assumptions in the calculations that require discussion. All of the

equations discussed in this section were taken from Shigley’s Mechanical Engineering

Design, 9th

Edition [3] unless otherwise specified.

1) The torque requirements for the drive motor were based on known equations for

ground vehicle dynamics [3]. The analysis was conducted by taking the variable

inputs and placing them sequentially into six key equations. These equations are

as follows:

31

Table 2.4 shows each variable involved in the calculations and its

corresponding value for this application.

Table 2.4. Variable Inputs for Driving Torque Calculations

Input Symbol Value

Gross Vehicle Weight GVW 240 lb

Weight on each drive wheel Wd 60 lb

Radius of wheel R 2 in

Desired top speed Vmax 1.76 ft/s (1.2 mph)

Desired acceleration time atime 2 sec.

Max incline angle α 0

Resistance factor RF 1.15

Frictional coefficient μ 0.4 (rubber and concrete)

Worst working surface f 0.015 (concrete, fair)

The gross vehicle weight was determined by making a rough weight

estimate for each component. The weight on each drive wheel is simply the total

weight divided by four, since there will be at least four wheels in contact with the

ground at all times. The radius of the wheel was chosen to be 2” because the

bottom of the drive motor would have come in contact with the ground had a

radius of 1.5” been chosen. 2” is the next standard wheel size that gives adequate

space. The estimated vehicle speed of 3mph was initially chosen based on a

reasonable estimate of an average person walking. It was later reduced to 1.2 mph

for the speed output of the drive motors that were selected. Although future designs

will potentially want to consider making the device faster, the prototype was concerned

only with getting a minimal amount of motion to prove functionality. The desired

acceleration time was selected to accelerate to 1.2 mph from a standstill in 2

32

seconds.17

The max incline angle was assumed to be zero since the demonstration

of the prototype will only occur on flat ground (or in the case of stairs, motion

parallel to the flat ground). Most of the time this device will be travelling on

concrete, so an intermediate value of “fair” was chosen. The value of “fair” falls

between “poor” and “good,” which are both measurements of the quality of the

traction on the concrete. “Fair” corresponds to a numerical value of 0.015, where

all values of concrete typically range between 0.01 and 0.02 [2]. The final

equation, which calculates the maximum tractive torque (MTT), was used as a

check to ensure that the wheel would not slip on the ground’s surface during

operation. If it could be shown that value of MTT is greater than the value of the

drive torque for one wheel (Tw), then the wheels would not slip. MTT was

calculated to be 48 in-lb, while the drive torque was calculated to be 26.1 in-lb.

Therefore, the wheels would not slip during operation.

2) The deflection in the drive motor mounting bracket was critical in determining the

thickness of material in the bracket required in order to keep the motor from

contacting the ground. The portion of the bracket parallel to the ground was

modeled as a cantilever beam with an end load and end moment. The end load is

simply the weight of the motor, while the end moment is caused by the fact that

the center of gravity of the motor is not in line with the end of the bracket. Figure

2.27 shows a diagram of the system.

17 The acceleration time was discovered to be much less during testing. This was because the motors spun

up to full speed without any motion control. Some reasonable estimate still had to be given for the sake of

the torque calculations, however.

33

Figure 2.27 - Diagram of Driving Motor Bracket Analysis

There were three key equations used to calculate the deflection. They are

shown as follows:

The first equation calculates the deflection at the tip of the bracket due to the

moment caused by the weight of the motor. The second equation calculates the deflection

due to the weight of the motor itself. These are summed up in the final equation to give

the total deflection caused by the driving motor.

34

The equations for deflection required several inputs, which are all

included in Table 2.5.

Table 2.5. Inputs for Drive Motor Mounting Bracket Deflection

Input Symbol Value

Width of plate b 3 in.

Length of plate L 2.5 in.

Young's Modulus of steel E 30e6 psi

Moment caused by motor weight M 14 in-lb

Weight of motor F 7 lb

Thickness of plate t 0.125 in.

The main parameter for these equations is the plate thickness. Thicknesses

of 1/16” and 1/8” were inputted into the equations to determine deflection. A

1/16” plate thickness would cause 0.0438” of vertical deflection at the tip. This is

acceptable considering that there is approximately 0.5” between the base of the

motor and the ground. Since 1/8” steel plating was already available, however,

this size was selected for the assembly. A 1/8” plate would result in 0.005” of

vertical deflection.

3) Deflection of the wheel shaft was determined by modeling the shaft as a simply

supported beam with a weight in the middle (equal to ¼ of the weight of the entire

device). The only equation required for this calculation was the equation for the

deflection of a simply supported beam with a center load. This equation is:

35

The inputs for this equation are listed in Table 2.6.

Table 2.6. Inputs for Deflection of Wheel Shaft

Input Symbol Value

Diameter of wheel shaft d 0.375 in.

Length of shaft between supports (bearings) L 4 in.

Young's Modulus of steel E 10.4e6 psi

Weight on each wheel F 60 lb

With a minimum18

shaft thickness of 0.375”, the resulting deflection was

0.008”. This is essentially negligible and would not result in contact between the

wheel and its housing.

4) The effective tension and desired pulley reduction would determine the correct

pulley and timing belt sizes. In order to determine these values, five equations had

to be used. These equations calculate the pitch diameters of each pulley, the

length of the timing belt, the torque seen by the motor after the pulley reduction,

and lastly the effective tension itself. These five equations are as follows [4]:

18 The shaft was stepped to account for the difference between the wheel’s hole and the bore diameter of

the bearings. The smaller of the two diameters was selected for calculations.

36

The required inputs for these equations are shown in Table 2.7.

Table 2.7. Inputs for Tension and Reduction of Timing Belt

Input Symbol Value

Center distance C 3.5 in.

L-series timing belt pitch p 0.375 in.

Number of teeth of driving pulley zp1 10

Number of teeth of driven pulley zp2 22

Belt width b 0.5 in.

Max required torque to steer Tmax 25.5 in-lb

Pulley ratio n 2.2

Max allowable tension for L-series belt T1all 100 in-lb

Using the center to center distance and calculated pitch diameters, the

required belt length was determined to be 13.0”. Using the driving torque, number

of teeth, and pulley geometry, the effective tension of the system was calculated

to be 19.4 lb. [4]. The maximum allowable tension in an L-series timing belt is

100 lb, so the selection of a 130L05019

timing belt was adequate. Slots were

included in the steering motor mounting bracket to allow for belt tensioning.

5) The required steering motor torque was determined by estimating the contact

patch radius and using it as the moment arm to determine the frictional torque. If

the applied torque from the motor could be shown to be greater than the frictional

torque, then steering would be successful. A few key assumptions were made for

these calculations. First, a test was done to determine the size and shape of the

contact patch. An epoxy patch was painted onto a section of the bottom wheel.

19 The first three digits “130” represent length to one decimal place. “L” stands for L-series. “050” stands

for 0.5” belt width. The pitch of this L-series belt is 0.375”.

37

The bottom of the wheel was pressed down against a piece of paper and then

lifted up. This was done several times to get a consistent contact patch size. The

assumption made here was that the contact patch was a circle, which doesn’t

diverge much from the actual epoxy mark itself.20

Figure 2.28 shows one of the

contact patch marks.

Figure 2.28 - Contact Patch of Drive Wheel

Second, the steering conditions assume that the vehicle is not also driving.

While this assumption will produce a larger amount of torque to overcome the

friction (since the required steering torque is less when the vehicle is in motion),

this assumption will produce the worst-case torque that the motor must surpass.

Third, the coefficient of friction between hard rubber and cement in dry

conditions is 0.85 [5].

20 In the event that the contact patch is not circular, one could calculate an average radius and use that for

the frictional torque equations.

38

There were two main equations used to calculate the steering torque. The

first equation is:

This equation gives the frictional force between the wheel and the ground. The

second equation calculates the frictional torque.

As shown in the second equation, the frictional torque is found by multiplying the

radius of the contact patch by the frictional force between the ground and the

wheel. The inputs needed for these equations are listed in Table 2.8.

Table 2.8. Inputs for Steering Motor Torque

Input Symbol Value

Coefficient of friction between dry concrete and hard rubber μ 0.85

Weight on each wheel w 60 lb

Radius of contact patch r 0.5 in.

The largest frictional torque for a wheel to overcome on cement was

determined to be 25.5 in-lb. A steering motor and gear reduction with at least this

much continuous torque would be required for full functionality.

6) The deflection of the steering motor mounting plate is critical because too much

bending would result in the edge with the steering motor to contact the top of the

drive motor. The system was modeled as a cantilever beam (the mounting plate)

with an end load (the motor), since worst case deflection would be seen at the

very tip of the mounting plate. The equation for the deflection of a cantilever

beam with an end load is:

39

The inputs required for this calculation are included in Table 2.9.

Table 2.9. Inputs for Deflection of Steering Motor Mounting Plate

Input Symbol Value

Weight of steering motor F 8 lb

Cantilever length L 4 in.

Young's Modulus of steel E 30e6 psi

Width of plate b 3 in.

Height (thickness) of plate h 0.125 in.

The results indicate that there will be at most 0.012” of deflection. This is

acceptable as it does not exceed the maximum allowable deflection of 0.88”,

which is based on the spacing between the plate and the drive motor.

7) The deflection of the horizontal support shaft (one in parallel with each horizontal

actuator) is important to determine the angle at which Pair 1 and Pair 4’s vertical

actuators contact the ground. The deflection was calculated in two stages. First,

each shaft was modeled as a cantilever beam with an end load equivalent to ¼ the

weight of the platform. The length of the beam is the maximum stroke length of

the horizontal actuators (18”). The deflection at the tip is the portion of the

deflection only due to the bending in the aluminum rods. In order to acquire the

actual deflection of the tip, the deflection and angle at the edge of the platform

would have to be added. There were four key equations necessary to calculate the

deflection and angle due to both the aluminum and the platform itself. These

equations are:

40

The inputs required to calculate the deflections and angles are shown in

Table 2.10.

Table 2.10. Inputs for Horizontal Support Shaft Deflection

Input Symbol Value

Force seen by each shaft F 60 lb

Moment arm of horizontal extension a 18 in.

Center distance between linear bearings L 6.88 in.

Total length of platform Lp 48 in.

Young's Modulus of aluminum E 10.4e6 psi

Young's Modulus of platform (plywood) Ep 1.01e6 psi

Outer diameter of support shaft do 1 in.

Inner diameter of support shaft di 0.5 in.

The total deflection, acquired by adding ya and yp together, comes out to

be 0.346”. This means that the vertical legs will be contacting the ground at an

angle equal to the sum of θa and θp, or 1.2°.

8) The torque allowed by the wheel shaft press fit was determined by calculating the

minimum interference fit such that the hard rubber wheel and solid aluminum

shaft would not slip relative to each other. The inner diameter of the hard rubber

wheel was known to be 0.404 in., and the diameter of the aluminum shaft could

41

be sized to allow for sufficient diametral interference. The allowable torque needs

to exceed the drive torque of 26.1 in-lb. Two equations, one for the pressure in the

fit and one for the allowable torque, were used. These are:

The inputs required for these two equations are shown in Table 2.11.

Table 2.11. Inputs for Wheel Shaft Press Fit

Input Symbol Value

Nominal diameter of wheel shaft d 0.401 in.

Outside diameter of wheel do 4 in.

Inside diameter of shaft di 0

Poisson’s Ratio for aluminum vi 0.333

Poisson’s Ratio for hard rubber vo 0.5

Young's Modulus for hard rubber Eo 14.5e3 psi

Young's Modulus for aluminum Ei 10.4e6 psi

Coefficient of friction between aluminum and hard rubber f 1

Length of interference fit L 1.6 in.

Minimum required torque21

Tall 25.5 in-lb

A minimum diametral interference fit of 0.00265” was calculated based on

these parameters. A diametral interference fit of 0.003” was chosen for ease of

machining.

21 The torque was known from previous calculations. Simultaneously solving both equations would give

solutions for p and δ.

42

9) The vertical actuators would have to be mounted to the vertical connector blocks

using the three screw holes near the tip. The tensile force that these screws would

see is equal to ¼ the weight of the platform minus the weight of each wheel

assembly. In order to include a safety factor, the tensile force was assumed to be

¼ of the weight of the device without subtracting the wheel assembly weight. The

allowable tensile stress in each screw (made of 18-8 stainless steel) is 42.1 ksi.

This means that each screw could not handle more tensile stress than this without

permanent set. Using basic equations to the calculate tensile stress in the bolts, an

acceptable bolt size could be determined. The two equations calculated the tensile

stress area as well as the tensile (axial) stress per bolt. These equations are:

The inputs required to perform these calculations are included in Table

2.12.

Table 2.12. Inputs for Actuator Mounting Screw Stress

Input Symbol Value

Max force applied on each leg F 60 lb

Minor diameter of #4-40 machine screw d 0.106 in.

Number of screws Ns 3

Based on the calculations each screw would only see 2.27 ksi, which falls

far below the maximum.

43

10) The bending stress of the platform was determined by treating the platform as a

simply supported beam with a center load (mostly due to the battery and weight of

the platform itself). With these conditions, the maximum bending moment occurs

at the center of the platform. The moment of inertia was determined by the section

properties of the solid model. The only required equations were those for the

maximum bending moment in a simply supported beam with a center load, as

well as the bending stress given that maximum bending moment. These equations

are:

The inputs required to perform these calculations are included in Table

2.13.

Table 2.13. Inputs for Bending Stress in Platform

Input

Symbol Value

Maximum stroke length of horizontal actuators h 18 in.

Weight load seen by platform w 240 lb

Length of platform Lp 48 in.

Area moment of inertia of platform Ip 46.125 in.4

Distance to neutral axis c 2 in.

With these numbers the maximum bending stress was calculated to be 219

psi. For plywood, bending stress should not exceed 1668 psi. Therefore, with the

44

plywood and 2x4 assembly, the wooden platform would suffice to take all of the

loads.

2.2.3. Component Selection

Most of the components in the design for both the vertical legs and frame were

machined. There were some, however, that had to be purchased directly. The following is

a list of the mechanical components that were purchased. All of these parts are also listed

in the bill of materials in Appendix C.

1) Driving motors

2) Steering motors

3) Timing belt pulley assemblies

4) Banded ball thrust bearings for wheel shaft

5) Flex couplings

6) Turntable bearings

7) Linear actuators (30” and 18” strokes)

8) Linear bearings

All of these components had at least one driving factor behind their purchase.

While meeting the specifications of the analysis was the most important factor for all the

components, cost ended up being a significant factor in several decision.

1) The biggest challenge in the selection of the motors for both driving and steering

was minimizing cost. Detailed research was conducted into several different

manufacturers and distributors of new brushed DC motors, and all of the ones that

met the requirements were prohibitively expensive. Because of this cost issue,

45

research was conducted on used motors. Eventually, feasible solutions for both

the driving and steering motors were found. In the case of the drive motor, a right-

angle gear motor was found that ran at 12V, drawing no more than 8A under a

loading output of 26 in-lb. The right-angle configuration simplified the design by

not requiring a worm gear or flexible shaft setup to account for the 90 degree

angle between the motor (mounted parallel to the leg) and the wheel shaft

(mounted perpendicular to the leg). Figure 2.29 shows the right angle drive motor.

Figure 2.29 - Right Angle Geared Motor for Driving22

This specific motor also runs at 100 RPM under design loading conditions. This

translates to 1.2 mph with a 4” diameter wheel. The cost of this motor was only

$60, as compared to costs of no less than $200 for new motors with similar

specifications.

2) The selection of the steering motor followed a very similar process as the driving

22 100 RPM 12 VDC Gearmotor. Surplus Center. http://www.surpluscenter.com/item.asp?item=5-

1649&catname=electric. 05 Aug. 2013.

http://www.surpluscenter.com/item.asp?item=5-1649&catname=electric

http://www.surpluscenter.com/item.asp?item=5-1649&catname=electric

46

motor. A motor with at least 12.75 in-lb of torque (which would become the

required 25.5 in-lb. with the 2:1 pulley ratio) was needed to overcome the contact

patch friction. The same distributor which sold the driving motors also sold inline

DC gear motors for only $50 that met the specifications. Figure 2.30 shows the

inline type that was purchased.

Figure 2.30 - Inline Geared Motor for Steering23

This inline geared motor runs at 20 rpm at the design loading conditions, which

translates to 0.152 revolutions per second (including the 2:1 pulley ratio).

Although the steering motors are rated at 13A for full load, the motors won’t

actually be operated under this condition. The timing belt ratio reduction would

allow the steering motor to draw approximately the same amperage as the driving

motor. This motor is rated for 16.75 in-lb of torque, so the minimum torque

requirement was met.

23 20 RPM 12 VDC Inline Gearmotor. Surplus Center. http://www.surpluscenter.com/item.asp?item=5-

1726&catname=electric. 05 Aug. 2013.

47

3) The main factors behind the choice of a timing belt to steer the leg were the

allowable tension and the spacing. The allowable tension of an L-series timing

belt with ½” width and 3/8” pitch is 100 lb. The calculated 19.4 lb transmitted

under this applications loading condition falls far below the maximum. Originally,

gears were going to be used to solve the problem of connecting the mechanical

source (the motor shaft) to the leg. The distance between the motor shaft and the

edge of the motor's gearbox, in addition to the short length of the shaft itself,

made gears an impractical solution. Instead of trying to cram the motor as close to

the leg as possible, an alternative approach was taken. By using a timing belt, any

length could be chosen to allow the motor to be situated in such a way as to not

interfere with the rotation of the leg.

4) As previously discussed, the choice of banded ball thrust bearings was related to

the potential thrust load on the wheels. It wasn’t necessary to consider other

bearings mainly due to cost, as the ones that were selected for this project were

only $6.

5) The selection of flex couplings involved the bore diameters, max parallel

misalignment, and transmitted torque. Since the wheel shaft size was determined

by the diameter of the caster wheel hole, the flex coupling bore had to be sized

according to a similar dimension. The shaft size at the wheel is 0.401”, so the

diameter that fits inside the coupling could be stepped down to the standard size

of 0.375”. The other bore, however, had to be sized based on the diameter of the

driving motor shaft of 7/16”. Max parallel misalignment of 0.005” was

determined to be within the max value for the chosen flex coupling based on the

48

alignment geometry in the solid model. Lastly, the coupling had to be able to

transmit up to 15 in-lb of torque without slipping. The chosen coupling was rated

for 31 in-lb of continuous one-directional torque. Figure 2.31 shows the selected

flex coupling.

Figure 2.31 - HCR112-14-12 Flex Coupling24

6) The turntable bearing was mostly selected for ease of assembly. It comes with

four bolt holes on the corners of each of the plates, allowing for easy assembly to

the sandwich plate on one side, and the steering motor mounting plate on the

other. Figure 2.32 shows the layout of the turntable bearing.

Figure 2.32 - Turntable Bearing25

24 “HCR112-16-10”, Helical Products Company, Inc. http://heli-cal.com/db/Products/. 14 Sep. 2013.

http://heli-cal.com/db/Products/

49

Given that the bearing could handle the compressive load on the leg (this specific

bearing is rated for 200 lb of compressive load), the only remaining concern

would be bending. Since the speed of the device is so slow, there would be no

substantial inertial force that would cause the bearing to bend. This speculation is

not adequate enough for a robust design, however, so testing was later conducted

to verify that any possible bending load the bearing could experience would not

damage the bearing.

7) While there were several different types of linear actuators that would have been

adequate for this project, this specific one was chosen mainly for its slenderness,

robust design, and mounting capabilities. Figure 2.33 shows the linear actuator.

Figure 2.33 - Firgelli Automation FA-400-L-12-30 Linear Actuator26

Two different types were required: 30” stroke for the vertical legs and 18” stroke

25 “Square Turntable”. McMaster-Carr. http://www.mcmaster.com/#lazy-susan-bearings/=nxrsbg. 05 Aug.

2013. 26

400 lb Force 30” Stroke. Firgelli Automations. http://firgelliauto.ca/400lbforce30inchstroke-p-71.html.

05 Aug. 2013.

http://www.mcmaster.com/#lazy-susan-bearings/=nxrsbg

http://firgelliauto.ca/400lbforce30inchstroke-p-71.html

50

for the horizontal extensions. These designs are exactly the same, the only

variation being their stroke lengths. Both exceed the load requirement of 60 lb by

being able to provide 400 lb of force. This type of actuator moves at 0.5”/sec.,

which is a reasonable speed for the motion of the device. While current draw is

rated for 10A peak, several tests were done under loading to determine that the

actual current draw during application would be between 3-5A. The mounting

hole at the tip of the actuator allows for easy integration with the rest of the

design. A bolt can hold the tip in place inside the tip's specially designed pillow

block.

8) The selection of the linear bearings required three considerations: diameter,

mounting configuration, and load capacity. The diameter was based on the

analysis done on the support shaft. Since a 1” diameter with 1/4” wall thickness

aluminum tubing was chosen, a linear bearing with a 1” inner diameter had to be

selected. For ease of assembly, the type of linear bearings with a mounting block

was chosen. Figure 2.34 shows the linear bearing with its built-in mounting block.

Figure 2.34 - Linear Bearing

51

The specific linear bearings chosen are not meant to handle large bending loads,

but for the application of this project the relatively small bending load did not

produce any problems in the ability of the bearings to actuate the support shaft.27

The load capacity of the bearings is 850 lb of dynamic load, which far exceeds the

maximum 217 lb vertical reaction load actually exerted on each bearing.

2.3. Mechatronics Design

The mechatronics design was guided heavily by the mechanical design. The

electronic components had to be able to control all twenty joints of motion while relying

on a local power supply. The design can be broken down into two subsections: hardware

and software. First a description of the hardware will be given based on the mechanical

components being controlled. Second, a description of the coding will be given based on

the drivers for each of the components. The mechatronics system can be grouped into two

areas:

Linear actuator control

Driving and steering motor control

2.3.1. Linear Actuator Control

The linear actuators chosen for the prototype each have two wire leads: power and

ground. Both the 18” and 30” stroke actuators run on 12VDC, drawing up to 10A at full

load (rated for 400 lb).28

Switching the leads alternates the actuator between extension

27 This was tested with the full assembly and under full loading conditions. The linear bearings actuated

smoothly without any problems. 28

Testing was done to measure the current draw at the design load. This is covered in 3.

52

and retraction.

Since the actuator needs to be able to switch directions, an H-bridge circuit had to

be used. An H-bridge is simply a circuit that acts as a polarity switcher, capable of

switching the leads of a motor from forward to reverse, as well as brake. A simple and

cost-effective solution was running the actuator off of a motor driver. The motor driver

circuit consists of 10 critical pin connections: 1) Motor VIN, 2) Motor GND, 3) Motor

Output 1, 4) Motor Output 2, 5) INA, 6) INB, 7) ENC/DIAG, 8) PWM, 9) 5V board

power, and 10) board ground.29

Motor VIN and GND are for the 12V, high current power input that will

power the linear actuator.

Motor Outputs 1 and 2 are where the two leads from the actuator will be

connected to receive the signal from the board.

INA and INB are the logic pins that will determine the direction of the

actuator (extend or retract).

ENC/DIAG is the pin for either an encoder or a position sensor to

determine the position of the linear actuator’s tip.30

PWM is the pin that controls the amount of voltage going to the actuator.

This controls the speed of actuation.

5V power and ground are the two connections the motor driver itself needs

in order to power the circuit to run the actuator.

29 This configuration is based on the Pololu Robotics and Electronics Dual VNH2SP30 Motor Driver

Carrier MDO3A. 30

While position sensors were not used for this project, it is still important to note that the boards driving

each of the components are capable of receiving signals from such sensors.

53

Since the actuators must all run in pairs, dual motor driver chips were selected.

Figure 2.35 shows the dual motor driver chip.

Figure 2.35 - Pololu VNH2SP30 Dual Motor Driver Board31

The dual motor driver chips are equipped with all the necessary pins in addition to

a second set of Motor Outputs, INAs and INBs, ENC/DIAG, and PWM. To reduce the

number of digital I/O pins required to operate the system, the INA and INB pins of the

two sets of horizontal actuators were soldered together. This way whenever a signal is

sent from the microcontroller to extend or retract, the actuators will move in the same

direction.

In order to run all of the linear actuators, six of the dual motor drivers will be

required. This means there will need to be 20 digital I/O and 12 PWM pins for the

microcontroller to control the linear actuators.32

2.3.2. Driving and Steering Motor Control

The driving and steering motors have conveniently similar specifications. The

31 “Pololu High-Current Dual Driver Carrier Board (VNH2SP30 or VNH3SP30) pinouts”. Pololu Robotics

and Electronics. http://www.pololu.com/catalog/product/708. 05 Aug. 2013. 32

Once encoders are added, there would also need to be 12 analog pins.

http://www.pololu.com/catalog/product/708

54

driving motor runs on 12VDC with 6A continuous current at full load. The steering

motor also runs at 12VDC with a specified peak current of 13A.33

The driving and steering motors would be controlled in a similar fashion as the

linear actuators. While there are enough digital I/O pins to operate each motor driver’s

INA and INB pins separately, the PWM signals for all of the steering and driving motors

had to be consolidated from 8 down to 2.34

The single motor driver chips used to control the steering and driving motors are

made to handle loads running at between 5.5-24V while drawing up to 12A continuous

current (30A peak). Both motors fit within this requirement, and therefore the single

motor driver was chosen. The single motor driver chips use the same pin setup as the

linear actuator’s dual motor driver. Figure 2.36 shows this layout.

Figure 2.36 - Pololu VNH5019 Single Motor Driver Board35

33 Because no additional manufacturing data was available, tests were conducted for the steering motor in

no-loading and design loading conditions. The continuous current draw for the steering motor at design

loading condition never exceeded 4A. 34

The Arduino Mega board has 15 pins which can be used as PWM. 35

“VNH5019 motor driver carrier, labeled top view”. Pololu Robotics and Electronics.

http://www.pololu.com/catalog/product/1451. 05 Aug. 2013.

55

The resulting motor driver setup for the steering and driving motors will require

16 digital I/O and 2 PWM pins.36

2.3.3. Microprocessor Selection

The two main considerations behind the microprocessor were cost and adequate

number of pins of the correct type. The Arduino Mega 2560 board, run with an

ATmega1280 microprocessor, has a sufficient number of pins of each type to be able to

control all the joints of motion in a somewhat limited fashion. Table 2.14 shows the

available pins, as well as the number of pins required for the prototype’s operation.

Figure 2.37 shows the assembled board.

Table 2.14. Arduino Mega 2560 Pin Usage Breakdown

Pin Type Required Quantity

Digital I/O 36 39

Analog Input 20 16

PWM 14 15

Figure 2.37 - Arduino Mega 256037

36 For position control, 8 analog pins would also be necessary.

37 “Arduino Mega 2560”. Arduino. http://arduino.cc/en/uploads/Main/ArduinoMega2560_R3_Front.jpg.

05 Aug. 2013.

http://arduino.cc/en/uploads/Main/ArduinoMega2560_R3_Front.jpg

56

While there are a sufficient number of I/O and PWM pins, the Arduino Mega

does not have enough analog inputs to provide for sensors for every joint. This is

acceptable as the scope of the project did not include using sensors. The next version of

the prototype will include sensors for the joints.38

The Arduino Mega can be operated on 12V, but since a laptop would be required

for manual control during testing, the Arduino was simply powered by a USB cable from

the computer.

2.3.4. Power Supplies

There would be two different power supplies on board the device. The first is a

8.4V, low current battery pack used to power the driver chips. The second is a 12V car

battery to power each of the components through the driver chips. Each of these power

supplies are on board per the specification. A car battery was selected for this application

for a few reasons. First, it provides 12VDC, which is the same voltage that all the

individual components need to be operated at. Second, it can provide a large current to

run the device. Third, it is rechargeable.

In order to calculate the necessary amp-hours for a desired amount of operation

time, some estimates had to be made regarding the usage of various components on the

device. The basic estimates were made based on a combination of product data sheets and

number of components.

38 The addition of sensors in future iterations is discussed in 4.

57

Table 2.15. Current Draw Requirements

Type Estimated Continuous

Current at Design Load (A)

Component

Quantity

Total Current

Draw (A)

Actuator 5 12 60

Steering Motor 6 4 24

Driving Motor 6 4 24

If all of the components were to draw their continuous current at design load

simultaneously, then the system would be drawing 108A. The car battery selected has 35

amp-hours, which means the entire system could be operating at full power for 19.4

minutes. This is, of course, an entirely unrealistic estimate since the components will

never all be running at one time. In fact, with the two different types of motion (climbing

stairs and regular driving), there will never be more than 8 components moving at once.

Table 2.16 demonstrates the different operating conditions and their power requirements.

Table 2.16. Current Draw for Various Conditions

Operating

Condition

Motors

On

Actuators

On

Total Current

Draw (A)

Available Time at

Full Current (min)

Climbing Stairs 2 8 52 40

Regular Driving 4 0 24 87

Both of these estimates are worst-case, since only certain stages use eight

actuators. Also, only two motors will be used if the device is driving straight. Only four

are needed during turning and driving. Worst case conditions show the device can run off

of a 35 amp-hr car battery for at least an hour, assuming mostly driving and not a lot of

obstacle or stair climbing. This is enough to show functionality.

58

2.3.5. Overall System Configuration

The Arduino board, batteries, motor drivers, terminal blocks, and surge protection

circuit will all be placed on top of the platform. Figure 2.38 shows a diagram of the

configuration of the electronic components.

Figure 2.38 - Electronic Components Assembly Diagram

Figure 2.38 demonstrates all of the critical connections between the major

components of the circuit. There were three modifications made from the actual setup for

the sake of diagram simplicity. First, the data pins were all lumped together as one

connection going to the microcontroller.39

Second, the high current power and ground

connections from the terminal blocks will each have their own wires, and will not be

39 This is, of course, not the actual configuration. Each PWM, INA, and INB pin will have its own separate

connection to an I/O pin on the microcontroller. Simply because of the number of connections required for

this circuit, placing them all in the diagram would only produce confusion rather than clarity.

59

lumped together as shown in the diagram. Lastly, the power connections coming from the

board terminal blocks will in fact come from two terminal blocks (power and ground)

rather than one.

2.3.6. Software

The purpose of the software is solely to control the motion of the device. Since

the coding is only focused on hardware control, using the Arduino language greatly

simplified the effort required to get functionality. The code accomplishes hardware

control by sending output signals from the MCU’s digital I/O and PWM pins. These

signals are sent via the hardwired connections between the Arduino Mega and the motor

driver chips. The outputs from the I/O pins handle the motor’s INA and INB logic.

The initialization of the software mainly involves declaring several variables to

correspond to the numbered I/O and PWM pins connected to each of the motor driver

chips. The naming scheme identifies the motor driver being referred to. An example is

“int Pair_2_SMotor_L_INA = 22;”. This stands for Pair 2, steering motor, left side40

,

INA pin. The initialization of the function variables, I/O pins, and PWM pins occurs

outside of the main function for global access. When later adaptations of the prototype

are made that include position control, it will no longer be feasible to keep the variable

declarations as global.

There are five variables which are important for the program’s functions. Table

2.17 lays out each variable’s name, assigned initialization value, and purpose.

40 The left or right side is determined from the view of someone sitting on the platform looking forward.

60

Table 2.17. Global Variable Declarations

Variable/Constant

Name

Initialization

Value Purpose

Input_A 0

This variable is the dummy variable used in

the motor driver functions in order to assign a

specific value of true "1" or false "0" to that

motor driver's INA pin. This, along with

Input_B, determines the direction the motor

will spin.

Input_B 0



specific value of true "1" or false "0" to that

motor driver's INA pin. This, along with

Input_A, determines the direction the motor

will spin.

motor_on 70

This is a constant that represents the PWM

value that will be assigned to a specific motor

driver’s PWM pin. Potential values range

from 0-255.

motor_off 0

This is a constant that represents the PWM

value that will be assigned to a specific motor

driver’s PWM pin. Potential values range

from 0-255.

P_W_M_value 0



PWM value to motor driver's PWM pin".

state 0 This variable is used to represent the state

that the program is in.

The structure of the software relies heavily on calling functions to complete

different jobs. The program consists of essentially one “task” which transitions from state

0 to state 1 at the beginning of the program. State 0 is the initialization state, while state 1

is the state in which a lengthy switch-case structure is used for specific commands. The

61

switch-case implements the usage of keyboard commands that are issued by the person

controlling the device. Each case involves one type of motion. For example, case ‘a’

causes vertical actuator Pair 1 to extend. Each case accomplishes its motion by calling

one or more of the program’s basic functions. Since the program is all manually

controlled, there was no need for a cooperative multitasking structure. There are a total of

three functions used: motorForward, motorBackward, and motorBrake. A description of

each function’s properties is included in Table 2.18.

Table 2.18. Function Properties and Descriptions

Function Name Input Variables Return

Value Description

motorForward( )

int Input_A, int

Input_B, int

P_W_M_value

void

This function sets Input_A high

and Input_B low. It also assigns

P_W_M_value the value of

motor_on.

motorBackward( )

int Input_A, int

Input_B, int

P_W_M_value

void

This function sets Input_A low

and Input_B high. It also

assigns P_W_M_value the

value of motor_on.

motorBrake( )

int Input_A, int

Input_B, int

P_W_M_value

void

This function sets Input_A low

and Input_B low. It also assigns

P_W_M_value the value of

motor_off.

The remaining portion of the code lays out the details of the switch case. The

details of this switch-case can be found in the actual code, located in Appendix F.

2.4. Manufacturing

Every component machined for the leg was made out of metal, whereas the frame

was constructed almost exclusively out of wood. While Appendix A contains all of the

62

part drawings required to machine these parts, there were several considerations that were

involved in the process.

The only special consideration for the frame base involved reversing the direction

of the actuator holes. Each hole has a slit in order to fit the linear actuator inside of it, and

one half of the base had the slots facing one direction, while the other half had them

facing the reverse direction. Figure 2.39 shows this hole layout.

Figure 2.39 - Frame Base with Opposite-Facing Actuator Slots

The reason for this configuration is that the legs needed to be aligned with each

other, so the center of the circles needed to match up. Also, in order to make the linear

actuator motors overhang on the outside of the platform, the directions of the slots had to

be reversed. The reason that the motors were chosen to face away from the platform was

63

so that planks could be used to support the actuator from moving up and down. Some

kind of fixture had to be created in order to keep the actuator body fixed, and simply

adding a vertical board to the frame with a hole to support the motor was the simplest and

most workable solution. Figure 2.40 shows this feature.

Figure 2.40 - Actuator Support

All of the wood pieces were fastened using wood screws. Since each screw was

used to fasten two pieces of wood together, a slightly larger hole was drilled in the first

piece before the screw was drilled in. This is so that the threads on the screw would

create a compressive force between the two pieces of wood. This would not occur nearly

as well if the larger hole wasn’t drilled first, since the screw head is no longer acting as a

clamp for its lower threads. The spacing between screws wasn’t exact, as each screw is

rated for well over the required loading condition. The screws were placed as a best

estimate to ensure a rigid structure.

Most of the parts for the leg required manual milling work, while some required

the lathe. All of the operations were done using standard machining methods, and no

special tooling was required.

64

3. TESTING

3.1. Plan and Objectives

Proving that each of the sequences of motion could be achieved required two

different sets of tests. First, the device had to be able to climb up a set of stairs. Second,

the device had to be able to climb over an arbitrarily shaped but reasonably sized

obstacle. Each of these sets of tests was done by using a laptop keyboard as the

controller, and initiating the correct commands to move the different joints of motion in

stages. This section provides some images of the stages accomplished in the two tests, as

well as any relevant discoveries made about the device by each of these motions.

3.2. Stair Climbing Sequence Tests

The first set of tests was done on the stair climbing sequence. Each of the stages

of this sequence were pre-determined by the sequence laid out in 2.1.4. A specially

fabricated wooden staircase with steps made of 11” x 36” x ¾” thick plywood sheet was

made to allow the device to perform the sequence. As previously discussed, the treads

were kept at their standard length of 11”, while the risers were half-scaled to 4”. The

staircase was placed at the base of Pair 1’s legs, at which point the climbing sequence

could commence. Because many of the stages could be replicated without any structural

issues whatsoever, the entirety of the images of each of the stages of the stair-climbing

sequence are presented in Appendix D.

65

The configuration of legs for this test placed the motorized pairs at Pairs 2 and 4.

Pairs 1 and 3 therefore had caster wheels. Figure 3.1 shows the transition from Stage 1 to

Stage 2.

Figure 3.1 - Stage 1 to Stage 2 Transition

This transition shows how Pair 2 must raise its legs before the platform can be

moved to the right to place Pair 2 on the first step. Initially, it was thought that simply by

retracting the horizontal actuators of Pair 1 that the platform would simply slide forward

on Pair 3’s casters. However, this assumption encounters two problems. First, Pair 1’s

legs also have casters, which means there is no locking mechanism in place to ensure that

the platform would not slide off the stairs while the horizontals are retracting. This early

discovery made it immediately clear that motorized legs on Pairs 1 and 3 would be far

superior to having them on Pairs 2 and 4.41

A motorized pair could lock the wheels in

41 This addresses the configuration concern raised in 2.1.3.

66

place, not allowing the platform to slide. The second problem encountered while trying to

accomplish Stage 2 was that the casters weren’t rotating freely enough to ensure that the

platform would move straight forward. The initial angle of the casters would determine

which direction they would slide. Fortunately, these two issues could be overcome by

simultaneously driving the wheels of Pair 4 forward while Pair 1’s horizontal actuators

were retracting. This worked very well to bring the platform forward and allow Pair 2’s

legs to be placed on the first step. The success of this motion is shown in Figure 3.2.

Figure 3.2 - Completion of Stage 2

Each of the remaining even stages (Stage 4, Stage 6, etc.) faced the same issue.

Each time the method of simultaneously retracting Pair 1’s horizontal actuators and

driving Pair 4’s wheels forward worked to bring the platform forward. Thus, up to and

including Stage 9 of the sequence, each stage was successful.

67

The largest difficulties were encountered after Stage 9. The completion of Stage 9

is shown in Figure 3.3.


In order to transition from Stage 9 to Stage 10, Pairs 2 and 3 must be lifted the

distance of one step to allow the platform to move to the right and place Pairs 2 and 3

back down on the next step. This requires the entire load of the platform to be carried by

Pairs 1 and 4. Up to this point, there had been little issue with placing the full load on

these pairs. However, because of a mechanical issue in the horizontal actuators, the

vertical actuators in Pairs 1 and 4 had a tendency to tilt left-right (into and out of the

page). This created an instability in the system as it was climbing, such that Stage 11

could not be attempted. Risk of crashing was simply too great. In order to even

68

accomplish Stage 10, the aid of a person was required to hold the back legs straight up.

Figure 3.4 shows the completed Stage 10.42


The problem with the horizontal actuators was caused by the fact that the tubes

inside the horizontal actuators were rotating. This was a completely unforeseen feature of

the tube actuators that had been purchased to accomplish the motion. This rotation, while

a minor concern for the actuators themselves, causes the four brackets that hold up Pairs

1 and 4 to tilt left-right. This instability is amplified the more the vertical as well as the

horizontal legs are extended. A solution to the tilting was found, however. A rigid steel

rod was inserted between the actuators of Pairs 1 and 4. This rod forced Pairs 1 and 4 to

42 The platforms weight was rested on Pairs 2 and 3 so that the device could be shown in the image without

the aid of someone to hold it in place.

69

remain mostly straight, thereby dodging the effects of tilting. Figure 3.5 shows the rod

inserted between the actuators.

Figure 3.5 - Rigid Rod Added to Reduce Tilting

In addition to the rigid rod, a different type of actuator, would have aided in

reducing tilting. A final evaluation of the stair climbing sequence tests demonstrates that

the device as designed can complete all stages.

3.3. Obstacle Climbing Sequence Tests

The second set of tests was performed by controlling the device as it traversed an

obstacle. One of the key aspects of the motion during this sequence is for the platform of

the device to stay stable and flat as it is carried over an obstacle. The beginning of the

sequence involved the vehicle rolling up to the obstacle with Pair 1’s vertical legs

70

retracted. The obstacle used for this test was 10” wide (left-right on the page) and 14”

tall. Figure 3.6 shows the initial position before beginning Stage 1.

Figure 3.6 - Vehicle in Position to Begin Stage 1

In order for the vehicle to begin the sequence, the horizontal actuators of Pair 1

must extend. This way Pair 1’s vertical legs can extend and take the front load of the

vehicle. Figure 3.7 shows the completion of Stage 1.


71

According to the original diagrams, Stage 2 involves lowering Pair 1 while raising

both Pairs 2 and 3. Because of the instability of Pair 1 and Pair 4’s legs while extended,

only Pair 2 was raised. This is acceptable until Pair 3’s legs must travel over the obstacle.

The test failed, however, while attempting to complete Stage 3a. Figure 3.8 shows an

image of the device at Stage 2, prior to failure.


When looking at Pair 1 in this image, it is clear that the same tilting issue

experienced in the first set of tests is also being experienced here. Even with the addition

of the rigid rod to reduce tilting, the twisting in the horizontal actuators themselves

caused enough of an issue to discontinue the testing at this point. An attempt was made to

commence Stage 3a, which involved using the drive motors of Pair 4 to drive the device

forward. Only about one second after the drive motors turned on, Pair 1’s legs completely

gave way and the vehicle crashed on its front-left corner. Nothing was damaged or

72

displaced during the crash, but it became clear that accomplishing the obstacle climbing

sequence would have to be done with a smaller obstacle.

It is important to note that many of the design features were verified during the

testing stages. The steering and driving both functioned adequately on all four legs. The

wooden supports for vertical actuators of Pairs 2 and 3 functioned properly to keep the

actuators straight up and down. The platform itself demonstrated more than enough

rigidity, even when it crashed. Lastly, all of the electronics worked as-designed, giving

full functionality for the steering, driving, and motion of the linear actuators.

Because the device lacked position control, the climbing sequences had to be

accomplished using manual control. Because exact positions could not be achieved, an

efficient way of surmounting this problem was by fine-tuning the PWM values of each of

the motor drivers to force each of the actuators to move at the same speed. Due to stiction

on the inside of the linear actuators, each of them were prone to move at different speeds.

By changing the PWM values so that each actuator in a pair matched the speed of the

other, the sequences required less manual tuning in between stages. This not only saved

time during the testing process, it also laid the grounds for additional features to be added

for the next version of the prototype.

In order to accomplish each stage of any given sequence, it is obvious that

position control be ultimately involved. This ensures that the device can stop at a specific

location, while also paving the way for the device to be pre-programmed with certain

sequences so that the driver would not have to control a device with so many joints of

motion. Tweaking the PWM values, however, also revealed the need for velocity control

73

in addition to position control. If all of the actuators can be ensured to move at the same

speed, the platform can be ensured to stay level. Adjusting PWM values is not a final

solution, but it did show that the platform could stay flat during the motion in each

sequence.

74

4. CONCLUSIONS AND RECOMMENDATIONS

This thesis began with several stated objectives, all of which were met with

varying degrees of success. As was already stated in this document, there are various

additions to this version of the prototype that will be made in the future, and these will be

discussed here as well. Figure 4.1 and Figure 4.2 show final images of the solid model

and actual prototype, which can be used for reference in the discussion of the objectives.

Figure 4.1 - Complete Solid Model of 8-Legged Robotic Transporter

75

Figure 4.2 - Complete 8-Legged Robotic Transporter

The first objective was for the device to be able to travel up and down a flight of

stairs. As the testing verified, the device was able to accomplish this through all stages

while being manually controlled. A sequence of commands issued from a computer

keyboard controlled the motion of the device. This allowed it to ascend the staircase in a

series of stages, which were identical to those laid out previously in this document. This

objective was not perfectly met, however, since the staircase that was used was scaled

down for cost purposes. A future design will need to be able to handle a full sized

staircase.

The second objective requires the device to be composed of 8 vertical legs, 4 of

which are fixed to the platform and can only move vertically, with another 4 that can

76

move horizontally as well as vertically. This goal was achieved by using 12 different

linear actuators – 8 vertical and 4 horizontal. The 4 horizontal actuators were shielded

from bending by placing a hollow aluminum shaft parallel to it, which took all of the

gravity load of the device. While all the actuators would typically move in pairs during

all modes of motion, both the mechanical hardware and the software allow each of the

actuators to move independently of each other. As discovered during the testing, the

horizontal actuators had a problem with rotation. This would eventually have to be fixed

by choosing a different type of actuators for the design.

The third objective was to place a weight on top of the platform to simulate the

transportation of cargo or persons. While the mechanical components for this prototype

were not designed to be able to carry a person, a 30 lb car battery was placed on the

platform to simulate the weight as well as perform the critical task of supplying power.

The fourth objective required that the device be powered with an on-board supply.

This came in three forms. First, a car battery was used to supply high-current power to

the components through the motor driver chips. Second, a separate 8.4V, low amperage

battery pack was used to power the chips. Third, the USB cable connected to the laptop

computer powered the Arduino Mega 2560. While this was not necessary, the cable was

required in order to send keyboard commands, making it redundant to supply power from

the 8.4V battery pack. If needed, however, the battery pack could serve this function.

The fifth objective states that the platform on the device stay level at all times.

Without position control, this was challenging. Ultimately, manual control was able to

bring the actuators close to, but not perfect, alignment. The main reason for this

discrepancy is that the linear actuators did not have linear encoders, making it impossible

77

to give feedback that could have ensured that all actuators were extended to the same

length. Since the actuators extend and retract at different speeds because of internal

stiction, the legs had to be independently adjusted to achieve a reasonable state of

levelness. Despite the setbacks of having no position control, the lack of encoders did not

stop the device from being able to sufficiently demonstrate all stages of motion in the

testing phase. It is important to note that the lack of levelness that resulted was not a

problem with the mechanical design, simply a problem with the system not containing

components necessary for position control. For scope and cost reasons, these components

were saved for a future iteration of the prototype.

Given the relative success of each of these five objectives, it is important that a

general conclusion be made. The results of the research done in this thesis indicate that

pursuing a further iteration of this device is warranted. None of the issues encountered

are insurmountable. That being said, there are still several problems that were

encountered that need to be addressed for future iterations of the design. While most of

these changes constitute additional features which are outside the scope of the objectives

of this thesis, there are a few that represent improvements of design decisions made for

this prototype.

There are several changes that could be implemented to enhance the current

design. First, there were several parts of the prototype that were simply too large. The 4

ft. x 3 ft. platform is bulky and could be downsized. Now that there is a prototype that

spells out what electronic components are required to make the device run, a future

iteration could use this information to fill the available space more efficiently. The 3 ft.

width could be reduced comfortably by 6-8 inches without causing any concern for either

78

tipping or having enough space for electronics. The objective here would be to reduce the

width such that the device could comfortably fit between the rails of most staircases. The

4 ft. length dimension could be reduced at the cost of offsetting the horizontal actuators.

The original reason 4 ft. was chosen was to allow for opposing horizontal actuators to be

collinear. If one were to reduce the 48 inches required in order to fit two 18-inch stroke

actuators, they would have to be placed side by side.43

Another option would be to place

the horizontal actuators on top of each other, so that overlap would not result in offsetting

the vertical actuators. Placing the actuators on top of each other would likely require that

a different type of linear actuator be chosen entirely, however. This is because the right

angle of the motor to the stroke makes it difficult to stack it on top of another of the same

type. An important note on this option is that if the actuator type were to change for the

vertical legs, several of the mechanical components would have to be completely

redesigned. However, if only the horizontal actuators were changed, very few mechanical

components would require redesign. As discussed previously, the usage of track actuators

would be far preferable to tube actuators. Track actuators would provide substantial

additional rigidity to avoid the issue of twisting that caused some of the later stages of the

testing to fail.

Another group of components that are simply too cumbersome are the drive and

steering motors. Many components were designed to accommodate such motors, which

were selected for cost reasons. A future iteration would want to include smaller motors, if

possible. This would allow the drive and steering portion of the prototype to shrink in

43 If the 18” stroke length was to increase, it would also be difficult to reduce the 4 ft. dimension.

79

size, as it would not be necessary to support such large and heavy components.

In order to advance to a second iteration of the design, there are a few key

additions that will need to be made. First, the device will eventually need to have at least

some (if not mostly) autonomous elements. The only way to achieve this is to incorporate

position control into the design. This will require not only selecting different driving and

steering motors, but also purchasing linear encoders that can track the stroke of the linear

actuators. The second critical element to autonomy is the selection of actuators that have

adequate stroke lengths to achieve stair climbing, overcoming obstacles, etc. While the

concepts tested in this prototype can serve as a guide for this, some dimensions will have

to be reexamined. It is possible that both the 30” and 18” stroke actuators will have to be

increased.44

Once position control and correct actuator stroke dimensions are

implemented, the software can be rewritten to include a position controller for each of the

joints of motion. It is very likely that with the onslaught of many additional signals,

namely ENC/DIAG and possibly analog current sensing, a different microcontroller

board will need to be chosen (or a second Arduino Mega 2560 that can communicate

with the first one). The Arduino Mega 2560 succeeds for this prototype because it has

enough digital I/O and PWM pins to control all of the motor drivers. It does not,

however, have enough analog inputs or ENC/DIAG pins to handle position control for all

the motors. In this case that a second board is not added and the Arduino Mega 2560 is

completely replaced, it is also possible that the coding platform will need to be changed

as well. Starting with the format of Arduino code written for this prototype, one could

44 It is certain that the 30” stroke will have to increase, but the 18” stroke depends on the final length of the

platform.

80

expand and create adequate code for a semi-autonomous device.

Another feature that could be added to simplify the hardware and increase the

aesthetic appeal is to control the motor drivers remotely. This would remove the need for

the excessive number of data wires that were required in order to make this prototype

function. While power would still have to be supplied by a hard line, all of the data

transmission could be accomplished with remote signals.

81

BIBLIOGRAPHY

[1] Goren Michael; Goren, Jeremy E. Stair-Climbing Human Transporter. U.S. Patent

7,246,671, July 24, 2007.

[2] “Drive Wheel Motor Torque Calculations”, EML2322L – MAE Design and

Manufacturing Laboratory. University of Florida. 03 Aug. 2013.

http://www2.mae.ufl.edu/designlab/motors/EML2322L%20Drive%20Wheel%20

Motor%20Torque%20Calculations.pdf

[3] McGraw Hill. Shigley’s Mechanical Engineering Design 9th

Edition. New York: The

McGraw-Hill Companies, Inc. 2011. Print.

[4] “Timing Belt Theory”, Gates Mectrol Inc., 2006.

http://www.thomasnet.com/white-papers/abstract/101106/timing-belt-theory.html

[5] Young and Freedman. University Physics 12th

Edition. San Francisco: Pearson

Education, Inc. 2008. Print.

http://www2.mae.ufl.edu/designlab/motors/EML2322L%20Drive%20Wheel%20Motor%20Torque%20Calculations.pdf

http://www2.mae.ufl.edu/designlab/motors/EML2322L%20Drive%20Wheel%20Motor%20Torque%20Calculations.pdf

http://www.thomasnet.com/white-papers/abstract/101106/timing-belt-theory.html

82

Appendix A: PART DRAWINGS

99

Appendix B: ENGINEERING CALCULATIONS

111

Appendix C: BILL OF MATERIALS

Table C.1. Bill of Materials

Component Description Quantity

Right angle gearmotor 100 RPM, 12VDC 4

Inline gearmotor 20 RPM, 12VDC 4

Driving motor bracket Connector plate 4

Flex coupling HCR112-14-12, 7/16" and 3/8" bores 4

Bearing pillow block Pillow block for thrust bearings 8

Banded ball thrust bearings 0.375" bore 8

Wheel shaft Shaft for drive wheel 4

Rubber wheel - 4

Caster rubber wheel assembly - 4

Caster plate - 4

Wheel housing Housing for drive wheel 4

Large pulley 22 teeth, 2.875" OD 4

Small pulley 10 teeth, 1.438" OD, with setscrew 4

Timing belt L-series, 130L050 4

Turntable sandwich plate - 4

Turntable bearing - 4

Steering motor mounting plate - 4

Aluminum tip pillow block

short

Block connecting actuator tip to

motorized leg assembly 4

Aluminum tip pillow block

long

Block connecting actuator tip to caster

leg assembly 4

30" stroke linear actuator - 8

18" stroke linear actuator - 4

Vertical connector block Connects 30" linear actuators to 18"

linear actuators and parallel support shaft 2

Vertical connector block

reverse

Connects 30" linear actuators to 18"

linear actuators and parallel support shaft 2

Linear pillow block bearing - 8

Parallel support shaft Supports horizontal actuators from

bending 4

Plywood base 36" x 48" x 3/4" 1

Plywood side supports 6" x 48" x 3/4" 2

112

Plywood back supports 5.5" x 48" x 3/4" 2

2x4 vertical actuator supports 28" height 4

2x4 base length supports 48" length 2

2x4 base width supports 36" length 2

Arduino Mega 2560 Microprocessor board 1

Double Motor Drivers Controls linear actuators 6

Single Motor Drivers Controls steer and drive motors 8

Battery pack Supplies board power 1

Car battery Supplies component power 1

23 slot ground terminal block Splits high amperage connections 2

15 slot ground terminal block Splits low amperage connections 2

50 amp switch - 1

40 amp fuse - 1

Fuse box - 1

22 gauge copper wire 300', red, black, and green 1

14 gauge copper wire, stranded 120', red and black, for high current

connections 1

8 gauge copper wire, stranded 6', red and black for battery terminal

connections 1

Wire staples Single nail with hook 33

Steel piping strap 6" segments 8

6-32 bolt, 1/4" length Connects steering mounting plate to

turntable bearing 16

6-32 hex nut Connects steering mounting plate to

turntable bearing 16

8-32 Threaded rod, 2.5" length Connects wheel housing, large pulley,

and turntable sandwich plate together 16

8-32 hex nut Fastens onto threaded rod 32

10-32 bolt, 3/4" length

Connects bearing pillow blocks to wheel

housing, and steer motor to mounting

plate

48

10-32 hex nut

Connects bearing pillow blocks to wheel

housing, and steer motor to mounting

plate

48

12mm bolt, 1 1/2" length

Connects actuators to tip pillow blocks,

and actuators to vertical connector

blocks

16

12mm hex nut

Connects actuators to tip pillow blocks,

and actuators to vertical connector

blocks

16

113

6mm bolt, 1" length Connects drive motor to bracket 12

5/16" bolt, 1 1/4" length Connects caster wheels to caster plate 4

5/16" bolt, 1" length Connects caster wheels to caster plate 12

5/16" hex nut Connects caster wheels to caster plate 16

5/8 ID nylon spacer, 7/16"

length

Separates caster plate from caster

assembly 16

1/4" bolt, 1 1/2" length Connects linear bearings to wood base 32

1/4" hex nut Connects linear bearings to wood base 32

8 GA wood screws, 1 1/2"

length Wooden platform assembly 74

4-40 machine screws, 0.62"

length Connects actuators to connector blocks 12

114

Appendix D: DATA SHEETS

117

Appendix E: SEQUENCE TESTING IMAGES

Figure E.1 - Stair Climbing Stage 1


118



119



120



121



122

Appendix F: SOFTWARE CODE

/*

Manual controller for an 8-legged robotic transporter

This code allows 20 different joints of motion to be controlled manually

by inputting keyboard commands.

*/

int Input_A = 0; // Function variable for INA pins

int Input_B = 0; // Function variable for INB pins

int P_W_M_value = 0; // Function variable for PWM pins

int motor_speed = 0; // Function variable for varying motor speed

int motor_1L = 160; // PWM value for Pair 1 actuator left side

int motor_1R = 160; // PWM value for Pair 1 actuator right side







int motor_5L = 255; // PWM value for front horizontal pair left side

int motor_5R = 90; // PWM value for front horizontal pair right side

int motor_6L = 120; // PWM value for back horizontal pair left side

int motor_6R = 120; // PWM value for back horizontal pair right side

int motor_on = 70; // Constant for PWM value assigned to drive and steer motors

int motor_off = 0; // Constant to turn off motors

int state = 0; // State variable for switch-case

//Pair 2, Left Side

int Pair_2_SMotor_L_INA = 52; // INA of the left steering motor for Pair 2

int Pair_2_SMotor_L_INB = 53; // INB of the left steering motor for Pair 2

int Pair_2_DMotor_L_INA = 14; // INA of the left drive motor for Pair 2

int Pair_2_DMotor_L_INB = 15; // INB of the left drive motor for Pair 2

//Pair 2, Right Side

int Pair_2_SMotor_R_INA = 16; // INA of the right steering motor for Pair 2

int Pair_2_SMotor_R_INB = 17; // INB of the right steering motor for Pair 2

int Pair_2_DMotor_R_INA = 18; // INA of the right drive motor for Pair 2

int Pair_2_DMotor_R_INB = 20; // INB of the right drive motor for Pair 2

//Pair 4, Left Side

int Pair_4_SMotor_L_INA = 22; // INA of the left steering motor for Pair 4

int Pair_4_SMotor_L_INB = 23; // INB of the left steering motor for Pair 4

int Pair_4_DMotor_L_INA = 24; // INA of the left drive motor for Pair 4

int Pair_4_DMotor_L_INB = 25; // INB of the left drive motor for Pair 4

//Pair 4, Right Side

int Pair_4_SMotor_R_INA = 26; // INA of the right steering motor for Pair 4

int Pair_4_SMotor_R_INB = 27; // INB of the right steering motor for Pair 4

int Pair_4_DMotor_R_INA = 28; // INA of the right drive motor for Pair 4

int Pair_4_DMotor_R_INB = 29; // INB of the right drive motor for Pair 4

//Digital I/O and PWM pin declarations

int SMotor_PWM = 2; // PWM of steering motors

int DMotor_PWM = 3; // PWM of driving motors

//Pair 1, Vertical Actuators

int Pair_1_VActuator_L_INA = 30; // INA of the vertical acutators, Pair 1, left side

int Pair_1_VActuator_L_INB = 31; // INB of the vertical acutators, Pair 1, left side

int Pair_1_VActuator_L_PWM = 4; // PWM of the vertical actuators, Pair 1, left side

int Pair_1_VActuator_R_INA = 32; // INA of the vertical acutators, Pair 1, right side

int Pair_1_VActuator_R_INB = 33; // INB of the vertical acutators, Pair 1, right side

int Pair_1_VActuator_R_PWM = 5; // PWM of the vertical actuators, Pair 1, right side






123

















//Front Horizontal Actuator Pair

int Front_HActuators_INA = 47; // INA of the front horizontal actuators

int Front_HActuators_INB = 44; // INB of the front horizontal actuators

int Front_HActuator_L_PWM = 45; // PWM of the front horizontal actuators, left side

int Front_HActuator_R_PWM = 46; // PWM of the front horizontal actuators, right side

//Back Horizontal Actuator Pair

int Back_HActuators_INA = 48; // INA of the back horizontal actuators

int Back_HActuators_INB = 49; // INB of the back horizontal actuators

int Back_HActuator_L_PWM = 12; // PWM of the back horizontal actuators, left side

int Back_HActuator_R_PWM = 13; // PWM of the back horizontal actuators, right side

// The setup routine runs once when you press reset

void setup()

{

Serial.begin(9600);

// Declares Driving and Steering motor pins as outputs:

pinMode(Pair_2_SMotor_L_INA, OUTPUT);

pinMode(Pair_2_SMotor_L_INB, OUTPUT);

pinMode(Pair_2_DMotor_L_INA, OUTPUT);

pinMode(Pair_2_DMotor_L_INB, OUTPUT);

pinMode(Pair_2_SMotor_R_INA, OUTPUT);

pinMode(Pair_2_SMotor_R_INB, OUTPUT);

pinMode(Pair_2_DMotor_R_INA, OUTPUT);

pinMode(Pair_2_DMotor_R_INB, OUTPUT);

pinMode(Pair_4_SMotor_L_INA, OUTPUT);

pinMode(Pair_4_SMotor_L_INB, OUTPUT);

pinMode(Pair_4_DMotor_L_INA, OUTPUT);

pinMode(Pair_4_DMotor_L_INB, OUTPUT);

pinMode(Pair_4_SMotor_R_INA, OUTPUT);

pinMode(Pair_4_SMotor_R_INB, OUTPUT);

pinMode(Pair_4_DMotor_R_INA, OUTPUT);

pinMode(Pair_4_DMotor_R_INB, OUTPUT);

pinMode(SMotor_PWM, OUTPUT);

pinMode(DMotor_PWM, OUTPUT);

// Declares Actuator pins as outputs:

pinMode(Pair_1_VActuator_L_INA, OUTPUT);

pinMode(Pair_1_VActuator_L_INB, OUTPUT);

pinMode(Pair_1_VActuator_L_PWM, OUTPUT);

pinMode(Pair_1_VActuator_R_INA, OUTPUT);

pinMode(Pair_1_VActuator_R_INB, OUTPUT);

pinMode(Pair_1_VActuator_R_PWM, OUTPUT);










124










pinMode(Front_HActuators_INA, OUTPUT);

pinMode(Front_HActuators_INB, OUTPUT);

pinMode(Front_HActuator_L_PWM, OUTPUT);

pinMode(Front_HActuator_R_PWM, OUTPUT);

pinMode(Back_HActuators_INA, OUTPUT);

pinMode(Back_HActuators_INB, OUTPUT);

pinMode(Back_HActuator_L_PWM, OUTPUT);

pinMode(Back_HActuator_R_PWM, OUTPUT);

// Default all pin outputs to low

analogWrite(SMotor_PWM, motor_off);

analogWrite(DMotor_PWM, motor_off);

digitalWrite(Pair_2_SMotor_L_INA, LOW);

digitalWrite(Pair_2_SMotor_L_INB, LOW);

digitalWrite(Pair_2_DMotor_L_INA, LOW);

digitalWrite(Pair_2_DMotor_L_INB, LOW);

digitalWrite(Pair_2_SMotor_R_INA, LOW);

digitalWrite(Pair_2_SMotor_R_INB, LOW);

digitalWrite(Pair_2_DMotor_R_INA, LOW);

digitalWrite(Pair_2_DMotor_R_INB, LOW);

digitalWrite(Pair_4_SMotor_L_INA, LOW);

digitalWrite(Pair_4_SMotor_L_INB, LOW);

digitalWrite(Pair_4_DMotor_L_INA, LOW);

digitalWrite(Pair_4_DMotor_L_INB, LOW);

digitalWrite(Pair_4_SMotor_R_INA, LOW);

digitalWrite(Pair_4_SMotor_R_INB, LOW);

digitalWrite(Pair_4_DMotor_R_INA, LOW);

digitalWrite(Pair_4_DMotor_R_INB, LOW);

digitalWrite(Pair_1_VActuator_L_INA, LOW);

digitalWrite(Pair_1_VActuator_L_INB, LOW);

analogWrite(Pair_1_VActuator_L_PWM, motor_off);

digitalWrite(Pair_1_VActuator_R_INA, LOW);

digitalWrite(Pair_1_VActuator_R_INB, LOW);

analogWrite(Pair_1_VActuator_R_PWM, motor_off);



















digitalWrite(Front_HActuators_INA, LOW);

digitalWrite(Front_HActuators_INB, LOW);

analogWrite(Front_HActuator_L_PWM, motor_off);

analogWrite(Front_HActuator_R_PWM, motor_off);

125

digitalWrite(Back_HActuators_INA, LOW);

digitalWrite(Back_HActuators_INB, LOW);

analogWrite(Back_HActuator_L_PWM, motor_off);

analogWrite(Back_HActuator_R_PWM, motor_off);

}

// Function for causing a motor to rotate in the "forward" direction

void motorForward(int Input_A, int Input_B, int P_W_M_value, int motor_speed)

{

digitalWrite(Input_A, HIGH); // Sets all motor power connections (red cable) to high

digitalWrite(Input_B, LOW); // Sets all motor ground connections (black cable) to low

analogWrite(P_W_M_value, motor_speed); // Updates the value of the PWM to the specified value

}

// Function for causing a motor to rotate in the "backward" direction

void motorBackward(int Input_A, int Input_B, int P_W_M_value, int motor_speed)

{

digitalWrite(Input_A, LOW); // Sets all motor power connections (red cable) to low

digitalWrite(Input_B, HIGH); // Sets all motor ground connections (black cable) to high

analogWrite(P_W_M_value, motor_speed);// Updates the value of the PWM to the specified value

}

// Function for causing a motor to stop

void motorBrake(int Input_A, int Input_B, int P_W_M_value)

{

digitalWrite(Input_A, LOW); // Sets all motor power connections (red cable) to low

digitalWrite(Input_B, LOW); // Sets all motor ground connections (black cable) to high

analogWrite(P_W_M_value, motor_off); // Updates the value of the PWM to zero

}

// The loop routine runs over and over again forever

void loop()

{

// send data only when you receive data:

if (Serial.available() > 0)

{

// read the incoming byte:

state = Serial.read();

// Switch-Case for motor selection goes here

switch(state)

{

case '1': // Case for extending Actuator Pair 1

{

motorForward(Pair_1_VActuator_L_INA, Pair_1_VActuator_L_INB, Pair_1_VActuator_L_PWM, motor_1L);

motorForward(Pair_1_VActuator_R_INA, Pair_1_VActuator_R_INB, Pair_1_VActuator_R_PWM, motor_1R);

}

break;

case '2': // Case for retracting Actuator Pair 1

{

motorBackward(Pair_1_VActuator_L_INA, Pair_1_VActuator_L_INB, Pair_1_VActuator_L_PWM, motor_1L);

motorBackward(Pair_1_VActuator_R_INA, Pair_1_VActuator_R_INB, Pair_1_VActuator_R_PWM, motor_1R);

}

break;


{



}

break;


{



}

break;


{



}

break;


{



}

126

break;


{



}

break;


{



}

break;

case '9': // Case for extending Actuator Pairs 2 and 3

{





}

break;

case '10': // Case for retracting Actuator Pairs 2 and 3

{





}

break;


{





}

break;


{





}

break;


{





}

break;


{





}

break;


{





}

break;


{





}

break;

case 'a': // Case for extending front horizontal actuators

127

{

motorForward(Front_HActuators_INA, Front_HActuators_INB, Front_HActuator_L_PWM, motor_5L);

motorForward(Front_HActuators_INA, Front_HActuators_INB, Front_HActuator_R_PWM, motor_5R);

}

break;

case 'b': // Case for retracting front horizontal actuators

{

motorBackward(Front_HActuators_INA, Front_HActuators_INB, Front_HActuator_L_PWM, motor_5L);

motorBackward(Front_HActuators_INA, Front_HActuators_INB, Front_HActuator_R_PWM, motor_5R);

}

break;

case 'c': // Case for extending back horizontal actuators

{

motorForward(Back_HActuators_INA, Back_HActuators_INB, Back_HActuator_L_PWM, motor_6L);

motorForward(Back_HActuators_INA, Back_HActuators_INB, Back_HActuator_R_PWM, motor_6R);

}

break;

case 'd': // Case for retracting back horizontal actuators

{

motorBackward(Back_HActuators_INA, Back_HActuators_INB, Back_HActuator_L_PWM, motor_6L);

motorBackward(Back_HActuators_INA, Back_HActuators_INB, Back_HActuator_R_PWM, motor_6R);

}

break;

case 'q': // Case for steering Pair 2 to the right

{

motorForward(Pair_2_SMotor_L_INA, Pair_2_SMotor_L_INB, SMotor_PWM, motor_on);

motorForward(Pair_2_SMotor_R_INA, Pair_2_SMotor_R_INB, SMotor_PWM, motor_on);

}

break;

case 'w': // Case for steering Pair 2 to the left

{

motorBackward(Pair_2_SMotor_L_INA, Pair_2_SMotor_L_INB, SMotor_PWM, motor_on);

motorBackward(Pair_2_SMotor_R_INA, Pair_2_SMotor_R_INB, SMotor_PWM, motor_on);

}

break;

case 'e': // Case for steering Pair 4 to the right

{

motorForward(Pair_4_SMotor_L_INA, Pair_4_SMotor_L_INB, SMotor_PWM, motor_on);

motorForward(Pair_4_SMotor_R_INA, Pair_4_SMotor_R_INB, SMotor_PWM, motor_on);

}

break;

case 'r': // Case for steering Pair 4 to the left

{

motorBackward(Pair_4_SMotor_L_INA, Pair_4_SMotor_L_INB, SMotor_PWM, motor_on);

motorBackward(Pair_4_SMotor_R_INA, Pair_4_SMotor_R_INB, SMotor_PWM, motor_on);

}

break;

case 't': // Case for driving Pair 2 forward

{

motorForward(Pair_2_DMotor_L_INA, Pair_2_DMotor_L_INB, DMotor_PWM, motor_on);

motorForward(Pair_2_DMotor_R_INA, Pair_2_DMotor_R_INB, DMotor_PWM, motor_on);

}

break;

case 'y': // Case for driving Pair 2 backward

{

motorBackward(Pair_2_DMotor_L_INA, Pair_2_DMotor_L_INB, DMotor_PWM, motor_on);

motorBackward(Pair_2_DMotor_R_INA, Pair_2_DMotor_R_INB, DMotor_PWM, motor_on);

}

break;

case 'u': // Case for driving Pair 4 forward

{

motorForward(Pair_4_DMotor_L_INA, Pair_4_DMotor_L_INB, DMotor_PWM, motor_on);

motorForward(Pair_4_DMotor_R_INA, Pair_4_DMotor_R_INB, DMotor_PWM, motor_on);

}

break;

case 'i': // Case for driving Pair 4 backward

{

motorBackward(Pair_4_DMotor_L_INA, Pair_4_DMotor_L_INB, DMotor_PWM, motor_on);

motorBackward(Pair_4_DMotor_R_INA, Pair_4_DMotor_R_INB, DMotor_PWM, motor_on);

}

break;

case 's': // Case for stopping all motors

{

motorBrake(Pair_1_VActuator_L_INA, Pair_1_VActuator_L_INB, Pair_1_VActuator_L_PWM);

motorBrake(Pair_1_VActuator_R_INA, Pair_1_VActuator_R_INB, Pair_1_VActuator_R_PWM);




128




motorBrake(Front_HActuators_INA, Front_HActuators_INB, Front_HActuator_L_PWM);

motorBrake(Front_HActuators_INA, Front_HActuators_INB, Front_HActuator_R_PWM);

motorBrake(Back_HActuators_INA, Back_HActuators_INB, Back_HActuator_L_PWM);

motorBrake(Back_HActuators_INA, Back_HActuators_INB, Back_HActuator_R_PWM);

motorBrake(Pair_2_DMotor_L_INA, Pair_2_DMotor_L_INB, DMotor_PWM);

motorBrake(Pair_2_DMotor_R_INA, Pair_2_DMotor_R_INB, DMotor_PWM);

motorBrake(Pair_4_DMotor_L_INA, Pair_4_DMotor_L_INB, DMotor_PWM);

motorBrake(Pair_4_DMotor_R_INA, Pair_4_DMotor_R_INB, DMotor_PWM);

motorBrake(Pair_2_SMotor_L_INA, Pair_2_SMotor_L_INB, SMotor_PWM);

motorBrake(Pair_2_SMotor_R_INA, Pair_2_SMotor_R_INB, SMotor_PWM);

motorBrake(Pair_4_SMotor_L_INA, Pair_4_SMotor_L_INB, SMotor_PWM);

motorBrake(Pair_4_SMotor_R_INA, Pair_4_SMotor_R_INB, SMotor_PWM);

}

break;

default:

Serial.print("Not a valid command.");

}

}

design and implementation of eight-legged robotic transporter

Documents